UKC

Who else has done this? It would seem an emotional wrench getting rid of all my CDs, or at least storing them in the attic, also selling or storing my top notch CD player (unfortunately no digital input to reuse its DAC) and buying an expensive streamer of similar or better quality, not to mention Tidal or similar subscription (life’s too short to rip all my CDs to an NAS but I’ll store some favourites on my phone). On the upside wide choice of music on demand and free up space in the lounge. 

In reply to kevin stephens:

I keep thinking about it but I can't bear to give up all the little booklets/lyrics & librettos/never before seen photos in sumptuous hand-tooled deluxe edition. I've also found it hard to love Kindle, despite the obvious ecological advantages of cutting down on all that paper. Have managed to adjust to enjoying my daily Guardian on the phone though.

In reply to kevin stephens: I can't remember the last time I bought a cd and I'm listening to more, different music than ever

In reply to kevin stephens:

I'm debating whether to flog my Marantz CD-67 Special Edition. I have got about half my CD collection on my NAS as FLAC but then ran out of steam. There all still in boxes since I moved house 3 years ago

I find myself to streaming off Spotify exclusively these days for better or worse.

In reply to kevin stephens:

I have bought a Roon Nucleus server, which has an internal SSD for digital files. I also have a subscription to Qobuz, which gives access to an almost infinite selection of music - I can read the music reviews in a newspaper and listen to the music that interests me immediately. The quality of the streamed music is more than acceptable.

The quality of the streamed files was improved significantly by buying an English Electric network switch, but CDs ripped to the internal SSD still sound better than the equivalent on Qobuz.

In reply to John2:

How can the switch possibly affect anything?

In reply to Rob Parsons:

Some sort of adaptive compression based on the bandwidth? 

All things being equal, a lossless digital file such as FLAC should sound identical to a CD, which would put any difference down to the DAC on the individual hardware.

I am, though, sceptical of people who claim to be able to hear the difference in such things.

In reply to planetmarshall:

I always thought digital and analogue electronic distortion isn't usually going to be very important compared to mostly mechanical distortion within the speaker (magnet surrounding a coil on compliant suspension connected to stiff diaphragm) plus various hi/mid/lo filters in a room with the listener breathing and birds tweeting or lorries passing outside. And a bit of tinnitus or ear wax!  

I still have the CDs, but I don't listen to them.

In reply to planetmarshall:

No network switch on the planet does that - all a network switch does is pass unmodified bits around.

This English Electric Network Switch sounds like an audiofool thing to me. However, let's hear the explanation ...

In reply to Rob Parsons:

No, the switch wouldn't, but a streaming device could modify the stream quality based on the network ping which could be affected by the switch. 

Hypothetically of course - I've no idea if Hifi component streaming devices actually do this, and you'd have to have a hell of a lot of internal network traffic for it to make a difference.

In reply to Rob Parsons:

The network switch reclocks the signal, resetting the time relationship between the digits. It also has noise reduction circuitry built in.

I'm not imagining things - the shop that I bought it from lent it to me for a few days so that I could try it in my own system.

In reply to John2:

From their website:

The 8Switch jitter measurements indicate an improved network signal performance of up to 90%.

Utter bullshit

In reply to Dark-Cloud:

Since the move of much of hifi technology to digital, I suspect many snake-oil salespeople are having to re-calibrate their marketing bullshit, since objective experiments showed that expensive speaker cable cannot provide a significant improvement in signal to noise ratio over a coathanger.

https://www.soundguys.com/cable-myths-reviving-the-coathanger-test-23553/

To each their own, though.

In reply to planetmarshall:

Let me break it to you that a speaker cable is a different product from a network switch.

In reply to kevin stephens: cost of a decent streamer/DAC = around 200 or more CDs. Tidal = 2 or 4 CDs per month. Hmmmm

In reply to John2:

Indeed, but the principle of selling a product based on producing an difference in sound quality that is below the level of human hearing, is not.

It's great that you're happy with it, but personally I'd want an objective demonstration before parting with £450.

In reply to planetmarshall:

Nah, everyone knows you need to spend all your money on a super duper fancy power cable to REALLY hear the difference.

Snake oil components in hifi and their die hard adherents are alway hilarious, personally I've always preferred spending the bulk of my budget on more music to play on my system.

Of course it's all non essential leisure purchases anyway so if people want to go crazy on cables let them, nobody's being forced to part with their cash.

In reply to kevin stephens:

I've been trying Spotify and a couple of things that let it down are that it doesn't have everything I want/have in my collection and finding stuff by browsing an artist is a pain for those with long careers.

In reply to J101:

Don’t forget for full performance it needs to be a directional cable 🙄

In reply to planetmarshall:

If you had troubled yourself to read my follow up post, you would have learned that the shop that I bought it from lent it to me for a few days before purchase so that I could try it in my own system.

In reply to John2:

Indeed not.  You're experiencing a placebo effect, which is an entirely different thing.  (And might even be worth the money, assuming you can spare it.)

In reply to John2:

And what difference did it make?

In reply to everyone:

This is really very tedious. I am describing what I have heard in my relatively high-end stereo system, and you are all airing your uninformed prejudices without having heard the product in question. If you want to know more go to the English Electric web site and read about the components included in the switch. I doubt that a single one of you could describe the method by which analogue sound in encoded into a digital signal.

In reply to John2:

Hi John, I've a decent system, albeit vinyl front end into valve amplification. I've previously dabbled in digital front end into solid state amplification so understand the processes.

If you're happy with your purchase and it makes your system sound better to you then that's all that counts.

Think you'll find that as often happens on forums people have veered off onto cables in hifi in general, which is a subject of constant intense debate (argument) on most hifi forums.

All the best with your listening.

In reply to kevin stephens:

Looks across at groaning shelves of CDs and hifi that hasn’t been powered on in a couple of weeks. 
Looks at phone and kitchen stereo used for hasty compressed streaming whilst making the tea.

No, course I won’t be selling up!

In reply to J101:

There’s a mirror image middle aged bloke argument going on on the Gear Page permanently over the relative merits of Germanium vs Silicon diodes in guitar overdrive circuits.

In reply to planetmarshall: looking at the link, cheap cable clips on the speakers indicate that the speakers in the test may not be likely to render faithful sound reproduction whatever the cables. 
For what it’s worth I’ve not experimented with different speaker cables, but higher quality inter-connectors made a massive difference to the sound of my system. No doubt someone who hasn’t heard it will post that I am a victim of confirmation bias. I put the difference down to how the properties of the cable affect operation of the output circuits of my CD player / input circuits of my amplifier 

In reply to paul_in_cumbria:

Except that it is at least just about feasible that the difference between germanium and silicon transistors used in a circuit deliberately designed to distort an analogue signal might actually be discernable to the human ear.

In reply to Removed User:

I've recently disconnected my marantz cd player from my old special edition amp. Ripped all my cd's a few years ago so never used it. They're still on the shelf though. Hardly ever stream the ripped files and mostly use Spotify now. 

I once played a few tracks switching between Spotify, Nas flac file and the cd. All through same amp and speakers -  couldn't tell much difference between them.

In reply to Jack: I understand that Spotify transmits at a lower bit rate than the more expensive sources like Tidal and this can make a big difference to sound quality, but I’d be interested in reports of first hand experience of this?

In reply to kevin stephens:

Analogue interconnect cables are a perfectly valid upgrade. All cables have capacitance, inductance and resistance and these will all affect the sound of an analogue signal. This effect is significant enough that it would be visible on an oscilloscope. 

I just looked at the English Electric 8switch. Utter woo. There is no way it's any different from any working switch. Either your previous switch was faulty or it's placebo effect. The website description is all complete meaningless drivel, here's just one example: "Switches require high speed power supplies so we could not opt for a linear power source due to their speed limitations. " Power supplies don't have "speed limitations"! There is no such thing as power supply "speed".

In reply to kevin stephens:

There's various places on the web that will play you the MP3 and the flac of bits of songs so you can test it yourself, I'm told. Google should find them. No personal experience - my music listening usually has various child-related noises to accompany it so I have assumed that currently I wouldn't ever get the opportunity to listen carefully enough to pick out the difference in practice, whether or not I could do so in test conditions.

In reply to kevin stephens:

Assuming you're talking about interconnecting cables carrying an analogue signal, it's significantly more likely that an upgrade has brought about an objectively perceptible change than it is in the case of an ethernet switch being swapped out for a more expensive one fitted with 'resonance-damping feet' instead of ordinary feet to prevent the digital bits from getting all jiggled about by 'unwanted mechanical vibrations'.

Though the only way to know for sure would be a double-blind A-B test, essentially the same way you eliminate placebo effects from a drug trial.

Nope, not a victim.  If you're enjoying the difference that you perceive (which may or may not be objectively 'real') you are at worst the beneficiary of confirmation bias - no more the 'victim' of anything than someone who feels better after taking a homeopathic remedy for a headache.

In reply to kevin stephens:

I think it's a better quality stream with Spotify premium than the free version. I stream it through a sonos connect, which supports up to cd quality. Comparing it with the actual cd played through the same system I was pressed to spot much difference. Same when I played the ripped flac file from a nas box. This was a few years ago - I may try it again and see, although the cd player has been put away.

In reply to deepsoup: 

 I also find it difficult to understand how digital cables can make a difference unless some are do bad as to lose bits. As far as my analogue cables go, several hours listening to the same Joni Mitchell with a number of borrowed cables demonstrated a clear difference, and the most expensive cables were not the best fit my system . 
 

Hope you’ve been getting out on the water lately?

In reply to kevin stephens:

I ripped all my CDs to my computer a few years ago and haven't bought a CD since. I was looking through our extensive dvd collection last year then realised I haven't had a dvd player for at least 4 years.

On the subject of quality between vinal, CD or streamed , 35 years working in industry means my hearing isn't good enough to tell the difference anyway. 

In reply to kevin stephens:

In a word: placebo!
Which isn't to say it isn't 'real' as such - listening to music is a profoundly subjective process after all.

Not nearly enough!  I got terrible pangs of Penrhyn Mawr envy from the photos of your exploits at the weekend.

In reply to kevin stephens:

I'm just hoping I can make it though to the end of my life still able to play my CD's at home and in my car. Switching from cassettes to CD's was stressful enough.

In reply to deepsoup: yes it was brilliant, I’m getting towards being self reliant in PM, two more coaching sessions with Roger over next few weeks, we’ll have to meet up on the water sometime 

In reply to Robert Durran:

Just for you:
youtube.com/watch?v=dSINO6MKtco&
youtube.com/watch?v=pujXTj4X_I4&

In reply to kevin stephens:

I like having the CDs, records and cassettes around, like other people like books or ornaments. If you're emotionally attached, keep them.

My local climbing wall uses Spotify. They frequently can't play my requests as they aren't available.

In reply to kevin stephens:

As a general reply to your original question:

1) If you enjoy listening to your existing music collection, why change anything?

2) Musicians will get much more money from you if you buy their CDs (particularly, if you buy them directly from the musicians) than they will if you stream their material. And good art needs to be paid for.

I got frustrated with playing CDs (and subsequently copied my entire collection to a local server using lossless compression, in a process which took me a couple of years) because I got sick of CD players skipping, and wanted a better way. But if you have a top notch player that shouldn't be an issue.

Clearly, streaming is where things are currently at but, whatever you do, please give some thought to my point 2) above.

In reply to kevin stephens:

We’ve moved a lot with work, and have dragged a load of crap around with us from house to house like an albatross around our necks. Mrs Paul in Sheffield had an interest in the Minimalists and Marie Kondo quite a while back, and had already really rationalised. 
I started with clothes, worked through my half of a library which I thought would be hard but when you apply the ‘when did I last wear/look at this?’, it’s surprisingly easy. Then the hard stuff. Applying the Five Watt World mantra of ‘1 guitar, 1 amp’, I’m down to 4 guitars and 2 amps, and I may take the last step soon.

All the climbing books have gone, except for current editions (and a BMC Froggatt guide which I bought in 1980 starting out). All the vinyl including Beatles etc my brother bought in the ‘60s. Finally, all the racks of CDs except for a handful of desert island disks to listen to when the power finally runs out, Led Zeppelin, Von Karajan Beethoven Symphonies, Megadeth, Bach Cello Suites and a few others. I *owned* a lot of stuff, but I never used it.

We are just about to make our final move, and as we’ll be cashing in and downsizing it’s really opened up our options. We had tons of stuff which we never touched and it’s been really liberating. Last night I though that I hadn’t heard the album ‘Live at the Hope and Anchor’ with The Stranglers, XTC etc since the sixth form common room in ‘78. Seconds later, it’s downloaded and playing on Apple Music.

Horror of horrors, the Linn Sondek, Quad amps and speakers are long gone, but I’m totally fine with it. I remembered that the best hifi I ever heard was a crappy mono reel to reel in the common room, playing Live at the Hope and Anchor.

In reply to deepsoup:

Yes, absolutely right. I’ve built pedals which can switch between Ge and Si signal paths and there is a difference, particularly how they clip a signal or react to overdrive in a multistage amplifier. However, the online flame wars argument is usually which is *better*, which is nonsense. Just like the valve vs transistor vs Digital modeling amp wars.

In reply to paul_in_cumbria:

I think I'll get that too with a bag on my head.

In reply to FactorXXX:

Thanks. The first one is timeless true genius. Works on many levels; it's how I feel about so much of life.

In reply to John2:

I jumped on Roon quite a while ago when there was a cheap life membership option. I'm very glad that I did - it's just great, isn't it? Will be even better when renovations are finished and I have a wired network again rather than trying to stream wirelessly. For me, Roon plus Tidal (and an old Unitiserve that I have assorted obscure CDs ripped on to) has been a huge change for the better. Mostly now via a Bluesound Powernode. I must confess to not wanting to go down a rabbit hole of expensive switches though...

But what do we do with our old CDs? There's a lot of non-recyclable plastic there....

b

In reply to kevin stephens:

My last remaining ways of playing a CD's are my Play Station and old laptop. Next generation of consoles won't play CD's IMO and then they go the way of the record player for most IMO. That includes blue rays and DVD's IMO. I give them 5 years or so until spinny things are dead

In reply to ben b:

Bird scarers on the allotment.

In reply to kevin stephens:

Switched 5 years ago…… get on with it! Lol 

In reply to kevin stephens:

I got rid of mine last year.  I'd ripped them about 10 years ago so they'd been gathering dust in the attic in two houses.  I have been pretty much exclusively used Spotify for years.  It was an emotional experience and I kept a big box of genuine rare stuff and albums I hope my kids will like when they get old enough to rummage through them.  I did sell anything with value to Music Magpie (got a pitifully small amount for the tens of thousands I must have spent).  I also gave the rest to the father in law for him to pick out anything he fancied and then he actually took them to the charity shop (and perhaps skip).  That step made it a bit easier - I didn't have to give them to a stranger.

Having got rid of them, I don't miss them.  The way I listen to music has change, my life has changed and sometimes you have to move forwards.

In reply to kevin stephens:

I was originally resistant to giving up purchased music but now happily use Spotify.  I've not got rid of the CDs though, they are in the loft for now.

In reply to kevin stephens:

When I first bought my CD transport (a long time ago) it had an optical connect option and a electrical connect option (both digital). I assumed they would sound either the same or that the optical would be better (given that it was being used in data transfer). The Optical cable sounded very audibly worse than the electrical connect. 

In reply to kevin stephens:

If anyone is using an old set up like myself it is well worth getting a DAC. I have it so that my spotify is playing on my laptop that is connected via USB to my DAC then RCA's to my Rotel Amp. I can then control the music with my Spotify mobile app. The quality of the music is brilliant.

If people want to look for high quality files to swap for their CD albums, Soulseek is very good.

I keep records though, they'll always be cool!

In reply to John2:

Hi John

"I doubt that a single one of you could describe the method by which analogue sound in encoded into a digital signal."

I'll bite!

I'm a sonar engineer by trade and unsurprisingly spend a lot of time working on analogue to digital sound. We also regularly use ethernet switches to pass data from subsea to surface. 

Never has the choice of switch used made a difference to our data quality. The sound is digitalized at source and decoded at the other end. the switch is simply a data bridge and as long as it meets ethernet standards then it is basically irrelevant.

What matters is the digitization and decoding at each end. The data packets for audio are tiny compared to other things so nothing particularly special is needed re the switch.  Ethernet its very stable and by its nature its not able to pick up interference or degrade sound quality. The data will come through exactly as it went in or it will come out as complete garbage that the decoder couldn't use in which case all audio would be wrecked (like the song would pause or skip a few seconds to the next good packet of data).

Regarding the timing relation issue you mentioned above, again that comes down to how you decode and process that data. its not how that data reaches the device. 

I could go into a more depth but i hope that makes some sense

first time my job has ever been relevant to UKC!

In reply to planetmarshall:

I used to read the James randi forums a lot and one of the frequent discusions were claims of massively improved sound quality by buying the latest gizmo, between those selling it and audio engineers saying that the claims were bull .

Sellers would defend their product to the hilt, but when challenged to do a double blind test to see if they could tell the difference, they seemed very reluctant

In reply to CurlyStevo:

Its going to be interesting when that happens. At the moment you can normally buy the game on disk for less than the digital download. My last purchase was Assassin's creed Valhalla, it was £60 on the Xbox store or £30 in Argos. 

Most of my games come from game pass as downloads for a subscription but for the price of the subscription its got to be a loss leader. Its basically the same as 1 full price game a year and just this year there have been 3 games so far that I would have paid full price for. 

In reply to John2:

You've knowingly plugged/put the network switch in circuit yourself so you know when it is being used.

The only actual way to tell the difference would be to have you blindfolded and then somebody to test the set up with and without the network switch and to see if you can tell the difference.

In reply to Dax H:

With PC games, I can go to CD Keys or G2A and get a game considerably cheaper than on steam/origin/blizzard etc. I got Dirt Rally 2 Game of the Year Edition for like £10 while on steam its £34. So still ways around paying full price. Don't know if its possible for Xbox and PlayStation though.

In reply to paul_the_northerner:

I'll give it a go too, off the top of my head. What you're essentially asking is how does a ADC work. Well, the analogue signal is sampled at a rate above the Nyquist limit. This ensures at least two samples per wavelength of the highest frequency you're trying to sample. In human hearing we generally consider the upper limit to be around 20kHz, so the sample rate should be at least 40kHz. In practice and due to some engineering considerations, 44.1kHz was chosen for the CD standard. That means the amplitude of the analogue waveform is measured 44,100 times a second, according to the clock of the sampler. Wobbles in the clock are considered 'jitter'. The more stable the clock the better. 

The accuracy with which the signal amplitude is measured is a result of the number of possible values. Again for the CD standard, a 16 bit word was chosen. This allows for 65,536 possible values (I confess I had to look up that number to be reminded....) The effect of more bits essentially pushes the noise floor down, giving greater dynamic range for the useful signal. 24 bit recording is common these days, even if the final result ends up at 16 bits.

 Anyway, you'll end up with a stream of 16bit words, 44,100 times a second. This can be stored on a CD or transmitted down a cable, and with checksums and other redundancy, can be perfectly recreated at the other end. Digital audio can put up with a lot of failure/errors in transmission and still be perfect on decoding due to this error protection/correction. It's an amazing system.

Lossy codecs, as used in much streaming audio, will analyse the audio and using psychoacoustic principals (such as frequency masking) throw out various bit of the signal where it thinks you won't notice. Once it's gone, it's gone forever. I listen in an amazing purpose built room, with a very high quality DAC, on speakers costing £12k/pair and can struggle to hear the difference between 320mbps and full-fat audio in a true A/B test. The idea that people can hear the difference in their sitting rooms is baffling to me.

Anyway, hopefully that helps

In reply to ChrisBrooke:

Sounds fantastic, but does your amplifier go up to eleven?

In reply to FactorXXX:

I never set it higher than about 6   My ears pay my mortgage, got to look after them.

In reply to John2:

I design AV systems for a living ( although admittedly not for the Audiophile market).... I'm fairly confident I could give it a go . It's also irrelevant here as the ADC stage has nothing to do with the switch - but anyway..  I was originally going to wade in with the WALOB approach but I'm intrigued. On the one hand knowledge of the TCP/IP protocol makes it difficult to resolve how the switch makes any difference BUT I don't necessarily disbelieve your, and many others, reports about the switch making noticeable improvements to their set ups. I'm not sure this can be explained by placebo alone. Could I ask - is your network switch racked with other audio equipment? What network cabling do you use - and most specifically do you use shielded cables -  either sftp or just ftp?

In reply to another poster:

"The website description is all complete meaningless drivel, here's just one example: "Switches require high speed power supplies so we could not opt for a linear power source due to their speed limitations. " Power supplies don't have "speed limitations"! There is no such thing as power supply "speed"."

I must admit that this does come across a bit woo but mostly because they are taking something that is based in science and then adding vague unhelpful statements. When they refer to 'speed' I assume they are referring to the PWM of the Switch Mode Power Supply that the unit uses (rather than a linear power supply. SMPS were always frowned upon for audio because cheap ones are terrible and noisy but there has been a move in some audio circles that good SMPS can outperform linears..... Anyway I mention this because at the moment the PSU is actually my best current guess as to why the 'Audio' network switch outperforms another one.

Anyway as a peace offering to all I offer this documentary (which I'm sure many of the Audiophiles here will have seen already)

youtube.com/watch?v=4b2IOOhJmxw&

As fascinating for the insight into what can be achieved from a lifetime's dedication to a single cause as it is for the geekery of the tech!

....and just so we don't digress from the thread... All of my CDs are in the loft (although my vinyl is still accessible). Why not buy an entry level DAC (or even see if a supplier will loan one) and give it a go? The up side to subscription services is you can unsubscribe if you decide they're not for you. You can probably even get trials of the various services to see how you get on with them.

In reply to NorthernGrit:

The problem is that none of those reports will be based on double-blind experiments.

In reply to Rob Parsons:

Perhaps. But it's also possible (even if remotely) that the switches are making a difference but with nothing to do with the network transport (which I am totally unconvinced could make a difference).

ie I'm thinking along the lines of EMI from PSUs or even grounding potential differences that are introduced onto the analogue portion of the signal chain either via proximity to the switch or due to shielded interconnects. Cheap switch has a cheap/noisy Power supply, 'Audio' switch has a decent power supply. Total guess. Perhaps a long shot and other options could be the case but dismissing out of hand isn't always a good starting point.

In reply to NorthernGrit:

But all such potential electrical effects should be measurable by instruments. If there was real data about this, why wouldn't the manufacturers involved make it very public?

In reply to NorthernGrit:

PS pretty obviously the network transport can make no difference at all: that music stream from Spotify will have traversed many switches before it finally reaches your 'audiophile' one ...

In reply to NorthernGrit:

I have a Cat7 cable running from my wireless router to my network switch, then the English Electric network switch comes with a high-quality Ethernet cable to go to my server. Are all Cat 7s sftp? I'm not sure.

The switch is racked with the rest of my hifi.

I have been assuming that the most effective component in the switch is the oscillator - to quote the EE web site -

'Powered by a customised TCXO (Temperature Compensation Crystal Oscillator), accuracy is 0.1ppm – far higher than normal crystals and higher than a normal TCXO, allowing the 8Switch to generate higher accuracy network signals, which in turn helps provide more stable music data transmission'

This is going to reduce jitter by providing the server with a perfectly synchronised data stream.

In reply to John2:

What device are you playing your music on?

It seems likely that anything playing audio is going to buffer the data before playing it which will eliminate any kind of network jitter.

In reply to John2:

That doesn't make any sense. The ethernet switch isn't sending out a constant stream of data which is then immediately converted to an analogue signal; rather, it's sending out discrete packets of data which are then fed to the buffer of the DAC, and finally converted to an analogue signal by that DAC. All the ethernet switch has to do is to faithfully transfer the data.

In reply to John2:

Yeah, sorry, not buying that. TCP is asynchronous. There are checksums on the packets etc. If it was UDP I might think there was an outside potential for it to make some difference to packet loss but not TCP.

Not sure we can thrash this out much further on this thread. Coupled with your setup description my best guess is that any objective differences are power related. Assume your switch and amp etc are on same power outlet? Anyway I'm really not being snarky when I say if the switch improves your experience then fill your boots.

In reply to kevin stephens:

I am probably being ignorant but I play my hundreds of cds via the player linked to my stereo system, and can do likewise with whatever I  get on Youtube or TV likewise and have huge pleasure of watching for example  Zimmermann play Beethoven no 3 full blast while watching him on the tv set

In reply to NorthernGrit:

I don't think there is any point in trying to have a technical conversation with someone who has read the EE web page and didn't immediately spot that it was all made up drivel.

If John2 likes his fancy switch then fine, but trying to justify this with woo is just going to annoy the likes of people who know what these words mean. 

In reply to NorthernGrit:

Hmm. I assume that the people who designed the switch included the oscillator for a reason. There is also the issue of noise reduction.

I am a consumer rather than a designer, and there is no question that the switch does improve sound quality.

In reply to John2:

If I'd paid > 10x over the odds for a switch I'd be pretty sure it improved the sound too. As well as improving the decor even though it was hidden in a corner.  

In reply to John2:

Who knows why they did it? Maybe they genuinely believed that since a switch needs a stable oscillator then a more stable one would make it better? Maybe they put it in knowing it was irrelevant but just so they would have a selling point? Maybe they didn't put it in at all and inside the box is a regular off the shelf switch?

But that blurb on the website only has two possible explanations: the person writing it didn't have a clue, or the person writing it was lying through their teeth. There are no other possibilities.

In reply to everyone :

I didn’t expect so much entertainment when I started this thread so thanks. The substance of the question seems to be more emotion than pragmatism. I am attached to owning my CDs and I also enjoy having and comparing different versions of some classical pieces. An appeal of streaming is access to new music, but these days you can buy a number of CDs for the cost of monthly subscription to a quality streaming service. I enjoy BBC Radio 3 and Radio 6 via my excellent FM/DAB tuner (Musical Fidelity A5 with valve output stage) as a route to discovering new music. Streaming internet radio would also offer other routes but part of the enjoyment of musical is not looking at a computer . In audiophile terms I guess my Musical Fidelity A5/M6 and PMC system is middle of the road, but a streamer to match would still cost a fair bit.

I will stick with what I have for now so thanks all.

In reply to kevin stephens:

You also need to factor in that at some point the software upgrades and support for any steamer you buy will stop, it's always put me off getting a dedicated streamer.

Plenty of options to put yourself one together from a raspberry pi base for not much money though, and ones other people have done themselves often come up on hifi classified or feebay etc as well.

In reply to kevin stephens:

I thought about it a few years ago, but then bought more CDs instead.  

I never play them, as they're all immediately ripped to FLAC files and stored on a NAS.  Yes, it was tedious to do this with my existing CDs but as compensation, I did find quite a few that I hadn't heard for years and had the pleasure of rediscovering.  

T.

In reply to kevin stephens:

My biggest issue with streaming is that the artist gets so little money from it. If you're a fan of band/artist and want them to continue making music for you enjoy then you really have to be prepared to buy something from them.

In reply to kevin stephens:

I ripped all my CDs. Started with those I'd collected up to 2010, with a concerted effort, and then ripped everything I've bought since; about 3000 now. Ripping is much quicker now, as metadata downloads easily, and rips proceed at about 25x real time.

I've tried streaming Qobuz, Amazon, Tidal and Deezer; they all give 3 month trials, which was handy during lockdown (despite having ~80k tracks ripped). I monitored the waveforms during playback, and found that much of it had been 'loudness wars' compressed, using soft limit compression to peg everything to FSD. It gets rather wearing after a while.

I also found that I just didn't use the services much, and that none of them provided selection or playback features anywhere near as good as tools like MusicBee; universally poor UIs, considering the money involved (cf MusicBee, written by one guy, in his spare time, and given away free).

I've just got the excellent LMS running under piCorePlayer, on an old v1.2 RPi B+. That gives you an excellent media server for about £15 for 2nd hand Pi and a 4GB uSD card. It will stream to Squeeze audio, UPnP/DLNA, Airplay & Chromecast end points, controlled via a nice, responsive web UI.

In reply to planetmarshall:

I post on a hifi website. I have learned to just sigh and ignore the most ridiculous claims. I have tried to educate, but some people want to believe. There's also the problem that people just don't understand the technology involved, and get emotionally invested in repeating their misunderstandings, and get upset when you try to correct them.

There are mechanisms by which both amplitude and time domain noise can couple to the DAC output, but a decent DAC should eliminate those mechanisms.

It's also fortunate that the website has an 'ignore member' function; the few people i have ignored are the loudest Dunning-Kruger types who think they are experts, but really don't understand. One tried to explain how a PC's DAC was involved in streaming via WiFi to an external DAC...

In reply to John2:

Because gullible people who don't understand the technology find it impressive. And pay over the odds for pointless stuff.

Like the others, I'm an electronic engineer. I built my first ADC when i was fourteen. I've worked on digital systems throughout my 35 year career, including streaming audio, video, networked systems, precision oscillators and digital RF modulators. A DAC clock doesn't need to be particularly accurate (the brain is good at ignoring that sort of thing, though just about any quartz oscillator will be more than accurate enough). What it does need, as does an RF synthesizer, is low phase noise ('jitter' in the digital audio world). That's why destination-clocked DACs are best, and why your switch oscillator is irrelevant; any jitter on the ethernet disappears in the FIFO buffer.

Oh crap. This is the sort of thing i try to avoid on the audio forum...

In reply to captain paranoia:

Or a locally-connected USB DAC...

In reply to planetmarshall:

By (un)happy coincidence I'm covering packet switching at this very moment as part of a course.  Wondering if you could explain this further as I'm not quite clear on how you'd use a network ping to modify quality of an audio file.  

I'm not especially up on ADC and DAC matters but wouldn't have thought something as crude as a network ping (which even on a local network takes several ms and can vary a fair bit) can usefully interact with something temporarily fine grained like an audio stream.  How does it work (in terms a 58 year old could understand)?

In reply to kevin stephens:

I got rid of all my vinyl (mostly collected from late 60s through mid 80s but with a few 50s 78s (and even a 16)  knicked from my parents).  I have to say I bitterly regret this, even though I have nothing to play them on.

Whenever these sorts of discussions come up my mind always turns to this, which I'm sure you've seen before:

youtube.com/watch?v=dSINO6MKtco&

In reply to Alyson:

I tend to make direct donations via their patreon accounts (or whatever they use) for stuff I like.  Or buy digital their audio files.

[ 3 posts to a thread in a row.  Boy am I trying to avoid getting to grips with this damn report assignment.  Must go away now and not come back for a bit. ]

In reply to Andy Gamisou:

I assume he means that if the ping had a very long return time that in theory it’s possible to write the app such that it would trigger a switch to requesting a lower quality stream to avoid risk of dropouts due to network lag.
 

It’s been a while since I looked but I thought some of the usual streaming apps had an “automatic” or “dynamic” setting in the “music quality” options, which I had always assumed was somehow assessing the amount of bandwidth available and requesting a file with a bit rate to match what was available.

In reply to captain paranoia:

What you are assuming is that you know how to measure all of the factors that affect perceived sound. In which case you would be better off working for a hifi manufacturer and capitalising on your knowledge.

I don't think that I have at any point claimed to know for certain how the network switch works - I have used locutions such as 'I assume' and 'presumably'.

All I have said with certainty is that I have tried Qozub both with and without the switch and I prefer it with.

In reply to Andy Gamisou:

Well you could be 58 year old DSP expert for all I know, however in basic terms - 

A CD has 41,000 16 bit samples per second, so requires a connection with at least 656kbps bandwidth to deliver that signal in realtime in uncompressed form.

Using the network ping (or other means to establish the bandwidth you have to your destination), you can get some kind of estimate of how much you need to compress that signal to deliver it in real time.

However...

This is obviously hypothetical. The bandwidth required to deliver a full-fat audio signal is tiny when compared to, say, a 4K H.265 video stream. You're much more likely to see this kind of thing if you're watching Netflix will simultaneously trying to upload something over a pitiful rural broadband connection...

In reply to kevin stephens:

If you can get your hands on a second hand Google Chromecast Audio, you can have CD quality streaming from Spotify, Tidal, whatever plugged direct to your amp via audio jack or via digital if you've a better DAC in your CD player for not much money. I bought mine new for £25 before they were discontinued and then bought two more for the kitchen and the bedroom as I was so impressed. Likely to be supported for a long time despite being discontinued as Google use the same cast protocols in their Google Home devices.

Edit: Just discovered there now exists the Audiocast M5, which is a straight rip off of the idea of the Chromecast (except it also works with DLNA) and even looks almost identical. Apparently the default software is crap but that's easily remedied.

In reply to planetmarshall:

A small nerd point, and one I didn’t make clear in my boring ADC description above: you’re looking at 16/44.1 times 2, as those samples are of course ‘per channel’ of a stereo source. 

In reply to kevin stephens:

I don’t have the technical knowledge of streaming audio that some on here do, but surely the network box just plops out data packets which may or may not be in the right order, and certainly aren’t time critical, which go in a big pot (buffer) in the decoder where they’re put in the right order, decoded, reclocked and DAC’d. It’s hard to imagine the network switch having anything to do with anything. Anyway, not trying to ‘pile on’….just wondering out loud. 

In reply to ChrisBrooke:

Exactly. The network transport can't possibly have anything to do with it.

As mentioned above, it is possible to imagine that electrical noise from the switch could somehow affect the end result. But without double-blind listening experiments, and/or actual measurements of electrical noise, all of this is meaningless speculation which is going nowhere. We might as well drop the subject.

In reply to captain paranoia:

What are you using for the ripping?

What are you using as the final target players? And which web UI are you referring to - that of LMS itself?

Thanks.

In reply to kevin stephens:

Hi Kevin. 

My views FWIW. I still have all my CDs. I occasionally play them, but they're all ripped to FLAC, live on my NAS and get played via a Raspberry Pi Streamer and DAC.

The biggest advantage of streaming for me (and I don't think anyone has mentioned this yet) is that streaming solutions like Volumio and Moode allow you to easily and cheaply implement digital room correction, which is well worth playing with. I also have all of my vinyl (and some I got from you!) and occasionally play it. 

Being a scientist by training, I do like the Audio Science Review forums, and if you do get the urge to upgrade anything, it's a good place to look for recommendations. 

Cheers

In reply to Paul Evans:

thanks Paul, that’s interesting. Although I have 2 very good Musical Fidelity DACs (tube output stages) they are locked in my CD Player and Tuner neither of which allows a separate digital input. I do have an unused Cambridge Dacmagic but it’s not a patch on the MF gear. I’ve decided that for the time being it’s not worth spending on a new DAC of comparable or better quality to enable streaming. I’m currently enjoying morning coffee with Radio 3 on FM

Also the NAS I bought recently, mainly for photo storage has packed up, need to try and arrange repair or replacement under warranty!

In reply to Paul Evans:

Can you briefly outline your entire set-up? What software are you running on the Pi? How does it see the NAS? Can the s/w use external streaming services as well as your own FLACs on the NAS, and, if so, which services? What DAC are you using? Etc.

In addition, what do you mean by 'digital room correction'?

Thanks.

In reply to kevin stephens:

I enjoy Spotify and Sonos for the convenience, but I still have a CD and vinyl collection and listen to both regularly. Spotify on my phone for use in the car mainly, and a Sonos speaker in the bedroom just for putting tunes on in a morning mainly, again streaming from my phone. The CDs and vinyl are in the lounge with a half decent stereo system. Personally, having your own collection of physical CDs/vinyl means you can easily browse through a collection of music you personally enjoy and have a connection to. Some of them I've had several decades and I keep coming back to them because they mean something to me. Many times I'll just scan the shelves and come across something I haven't listened to in years, but remember that I love it and stick it on. It's also a great way of sharing music you enjoy with others, putting things on when friends are over, and they can browse your collection too and find talking points etc. There's just so much more to it than a collection of playlists on Spotify. All have their place, but I'm never binning or storing away my CDs or vinyl.

In reply to Abu777:

I've moved to entirely streaming and my music collection on FLAC/MP3, purely because I never felt that much of an emotional connection with playing a CD and I'm still pretty good at listening to an album from start to finish rather than a playlist or shuffle.

If I end up in a larger house (baby has happened) then I think I'd get a decent turntable and start collecting my favourite albums on vinyl, which has always had more of an emotional / ritual element to it over any other ways of playing music for me. The simple act of having to get up, put the needle down, turn it over halfway and repeat seems to have me sit and be focussed on listening more consistently than flicking on a stream with my thumb from my phone as I'm chopping onions.

I'm actually not an evangelist about the added noise and 2nd harmonic distortion, sorry "warmth", that vinyl has over digital - it sounds good on some things and detracts from others in my experience, but it does sound like nostalgia, which is pleasant on the right record.

In reply to Rob Parsons:

I'm not Paul Evans, but I'll give this a go. 

When loudspeakers are placed in a room their acoustic output, where it meets your ears, will be coloured by the acoustic character of the space they're in.

Imagine a pair of headphones: there's a straight-forward connection between the driver and your ear. However, with speakers in a room the sound from each speaker is going to both of your ears directly, and also indirectly as it reflects off every surface. These reflections will combine to create peaks and troughs at different frequencies due to how each wavelength interacts with the physical dimensions of the room. i.e. Some frequencies will be a bit louder, some a bit quieter. You'll also get various comb-filtering effects and phase issues: the summation of all the reflections and how they combine with the direct signal occurs in the frequency and phase domains.  This is how the room 'colours' the sound. We're all instinctively used to this as we all listen in less than ideal spaces - even the professionals - but some rooms will sound better than others, intentionally or otherwise....

One way to deal with this is to measure the response across the frequency spectrum at your listening position and use an inverse EQ curve to 'counteract' the inaccuracies. The goal being a flat, or linear, response in the listening position. The simplest DRC systems will do this. At the practical level you need a quality measurement microphone to record the response (by playing some white or pink noise or a frequency sweep for example), then let the computer/plugin do the DSP to create the inverse EQ curve.

In practice I'e never heard an EQ system I like but people do seem to find them useful. The only system I like, and which we use at work is from Trinnov. That has a four-pronged measurement mic and is thus able to correct in the phase domain also. The results are truly astonishing. Even in a very good room you can squeeze out that last 5% of perfection. The phase correction is the most useful part - your phantom images become pinpoint accurate (e.g. a kickdrum/vocal/bass guitar that is meant to be central comes from a point right between the speakers, not just roughly somewhere between the speakers. Unfortunately the price is also astonishing.

In reply to Rob Parsons:

I've used Exact Audio Copy since i started. It tries very hard to get a bit perfect rip, and now integrates with AccurateRip to check it does. Any minor rip errors can he corrected with CueTools.

But MusicBee also does a good job of ripping, and also integrates with AccurateRip.

LMS is a headless server; it's web only. I use the Material Skin UI plugin; one of the many useful LMS plugins.

Endpoints? Lots...

Denon Ceol Piccolo UPnP/DLNA network receiver

Chromecast Audio

Squeezelite app running on PCs.

Squeezelite running on the Pi or other host, to USB DAC, or HAT DAC.

BubbleUPnP app on Android.

piCorePlayer, as the name suggests, is actually a version of Tiny Core Linux for the Pi, hosting Squeezelite. LMS is a bolt-on function... So you can make a piCorePlayer Squeezelite renderer with Pi Zero W, a tiny uSD card (the default partition for pCP is just 64M), and a USB headphone dongle DAC; the Apple dongle is very cheap, and really rather good.

In reply to Rob Parsons:

Hi Rob. 

I'll give it a go. Chris has explained DRC in more detail than I was going to (cheers Chris!) so I'll pass on that bit.

Currently the RPi (3b+) is running Volumio, but I'm planning to move back to Moode when I get some time, since the DRC looks better. Both are free. Volumio and Moode can see the NAS via wired or wireless LAN (I use wired) using (I think...) UPnP. You need to have the music in a music share on the NAS.

The RPi feeds a separate DAC, I have an Allo DigiOne HAT feeding SPDIF to the DAC, as the USB signal from the RPi 3 isn't that clean, but I'd ditch this if I was using an RPi 4.  Volumio and Moode both support the usual streaming suspects (check their websites for more detail), but I don't actually subscribe to any (odd I know). 

The DAC is a Khadas Tone Board which audio science review gave very high marks at the time, not bad for £100. Fed from a home made linear PSU. 

The DAC feeds a late 1980s Cyrus One (recapped and had to replace one of the preamp voltage regs). And that feeds some Monitor Audio BR5s.  

I also have vinyl and CD player, but I think your questions were more about the streaming solution. 

Hope that helps, let me know if you need any more info. 

Paul

In reply to John2:

That's the argument used by those who don't understand how these systems work, and where the causation barriers are (the FIFO I mentioned above is the most relevant barrier here).

Whereas I do know how a network switch works, and how digital streamed audio works (I've implemented digital streamed analogue systems). I know the potential degradations that are present.

The only one that can have an effect on the analogue audio in a destination-clocked streaming system is coupling of electrical noise from the digital interface to the analogue audio. That should be eliminated for all practical purposes by the digital interface of the DAC. But marginal gains...

So, if your fancy network switch is electrically cleaner than a bog-standard switch, and your DAC has a poor digital interface, it may improve the electrical noise coupling.

Your earlier post alluded to change of codec in bandwidth limited bearers, using QoS measures. This does happen in internet video streaming services; iPlayer certainly does this. Bluetooth audio protocols do this. But if you have bandwidth so limited it can't support even the highest resolution digital audio, you need to address your network issues; something is very wrong (in the context of modern broadband). If you're having these troubles streaming locally to your home network, something is very, very wrong.

In reply to Paul Evans:

They can access DLNA servers, but it's probably better to get them to connect to the NAS file server via CIFS/SMB, and create their own media database. Volumio will give much better gapless playback using that method.

In reply to kevin stephens:

I think I'm backwards -- I actually bought a few CDs recently, first time in years - so that when driving on trips I could listen to music regardless, of being on a motorway with great signal or a rural area with no data

Edit: Because you lot are nosy   they were Doors - LA Woman, Ram Jam - Ram Jam, Deep Purple - In Rock, Funkadelic - Maggot Brain

In reply to ChrisBrooke, Paul  Evans, captain paranoia:

Thanks for those various responses.

FWIW I use LMS on a local Linux server, feeding various Squeezeboxes in various rooms, and controlled in various ways (LMS UI; phone app; or physical remote controls.) And I rip to FLAC on the Linux server using cd paranoia (old - but it still works.) My FLAC collection is mirrored in various ways; should all those go tits up (which I hope never happens) I still have the CDs themselves as the ultimate backup.This system has served me very well for ten or so years but at some stage various parts of it will break so I am always interested to hear what others do.

For listening in the car I transcode to MP3 and then use whatever device the car deals with most easily (which right now is a USB stick.)

In reply to captain paranoia:

'Your earlier post alluded to change of codec in bandwidth limited bearers, using QoS measures'

It did? Now you really are puzzling me.

I have to say, the import of your posts seems to be that in order to enjoy a piece of hifi equipment you need to understand how it works. I'm not sure why you think the bandwidth of my network is limited.

As you say, marginal gains. They add up, as the British track cycling team demonstrated.

In reply to John2:

Oh, there's me thinking the "sum of marginal gains" theory was an obviously flimsy cover story first concocted by team Sky to explain their performance edge while the team doc completely coincidentally prescribed performance enhancing drugs to "close relatives" of the team.

Silly me..

In reply to John2:

I think the Captain may actually be referring to planetmarshall and others' posts, in terms of the compression point.

I don't read it as "you need to understand how it works to enjoy it", more "the reason you're enjoying it has very little to do with what the manufacturer claims it's doing" (i.e. reclocking a non time dependent signal doesn't do anything much, so perceived benefits potentially come from some blend of less electrical noise and placebo). 

In reply to AJM:

Well he did use the words, 'your earlier post' in a post replying to me.

In reply to John2:

Apologies, you didn't; it arose in the ensuing discussion.

No, you can enjoy it without understanding anything about it.

But if you try to justify why a magic box makes it sound nicer ("Powered by a customised TCXO (Temperature Compensation Crystal Oscillator), accuracy is 0.1ppm", "The network switch reclocks the signal, resetting the time relationship between the digits", "This is going to reduce jitter by providing the server with a perfectly synchronised data stream"), then you need to be able to explain why it makes it sound nicer, which does require understanding. If you want to improve your audio system, it's worth having a reasonable understanding so you know where to spend your money effectively.

The Ethernet interface is separated from the DAC by at least five causation barriers:

1. Ethernet physical layer protocol
2. Ethernet transport protocol (be that TCP/IP or UDP)
3. Streaming protocol (Squeeze,  UPnP, Airplay, etc), carrying CODEC audio packets over the transport protocol
4. Decoding of the CODEC packets(FLAC, MP3, etc) into PCM by some processor
5. Buffering of PCM samples into at least one FIFO buffer

Each of those barriers completely separates the network jitter from the DAC audio; there is no causal relationship between jitter (short-term data timing variation) on the Ethernet, and the DAC input samples, let alone the DAC audio output. The only place jitter matters is in the DAC sample clock; that is nothing to do with the Ethernet data clocking.

The Ethernet interface may carry electrical noise from the switch into the network streaming endpoint, as I said, and a badly designed unit might allow that electrical noise to get into the audio. But there is a lot of digital processing going on in that network endpoint, which also generates electrical noise; they've got to deal with that, too.

We can measure noise in DACs quite easily; the SNR of a well-engineered DAC can be way below the threshold of human hearing, and way below the noise floor of any listening room.

In reply to John2:

Yes, I read that too - I was merely trying to provide some clarity, since you said you were puzzled

In reply to captain paranoia:

I'm not trying to justify anything. I am reading the manufacturer's blurb (not a bad place to start) and attempting to use that to reply to those who mock the idea that a network switch might improve sound quality. I am quite happy to be corrected by those who point out that ethernet is a packet switching system.

As for the noise floor of the listening room, that's an irrelevance - a less noisy signal will sound better and less hashy in any listening room, whatever the noise floor.

In reply to AJM:

You know, many of the comments on this thread remind me with a conversation I had with a certain Rockfax guidebook writer in the south of France (you were present too, I think) who despite being a teetotaller insisted that more expensive wines were just a marketing ploy that tasted no better than cheaper wines.

In reply to John2:

Hehehe!

In this case, isn't it more like saying a bottle of wine tastes better if the corkscrew is a walnut handled one rather than having a plastic handle?

In reply to John2: 

Bet I could .  Bit too much seismic processing here

In reply to AJM:

Like the teetotal guidebook writer, you have never heard my stereo system.

In reply to John2:

Of course not.

There’s a part of stereo listening, like wine drinking, that’s subjective - some people won’t appreciate expensive wine more than cheap wine in the same way that some people will be perfectly happy with a mono Bluetooth speaker and mp3 streams and others would refuse to subject their ears to anything less than hi-res through a 10k dac.

But at the same time I tend to think there’s at least some objectivity in stereo (is there a universally accepted definition of “better” when it comes to taste? I don’t know). Is anyone for example going to argue that all else being equal a higher background noise level is better - a room with an air con unit running next door or next to a busy road for example. And some of it is science - it’s difficult to argue the physics in terms of the impact of the room in terms of how it changes what reaches your ear versus what comes out of the speakers. And therefore there’s at least some validity to the “should this make a difference” question.
 

Personal view then, from what I know of how switches work and how I know dacs work:

- I find it hard to believe the reclocking does very much - if nothing else I find it pretty hard to imagine the packets are arriving in real time, given most services probably buffer some music to avoid dropouts. If it’s going into a buffer somewhere, then it doesn’t matter if the last leg of its journey was precisely timed.

- I can buy the conceptual idea of electrical noise being a thing. I don’t have a 100% settled view on whether it’s a big thing or not but I tend to think a lot of other things would have to be optimised before it became the weakest link just because it’s several steps back down the chain

- it could be a genuine flaw in the kit you had previously that means otherwise minor improvements against normal kit become more major

- placebo is a well established enough phenomenon that it’s impossible to discount.

If you think it sounds better then that’s great. You’ve put some money into something that’s improved your listening experience. It doesn’t really matter why, as others have said. 

In reply to John2:

It isn't, because they are trying to sell you unnecessary stuff. There are loads of duplicitous bastards out there, trying to sell you stuff you don't need.

That's all I am trying to do; to educate. To help you (and anyone reading) understand how these systems work, so you won't be fooled by the duplicitous bastards.

Exhibit 1: is this a real site, or a parody:

http://www.machinadynamica.com/

In reply to AJM:

Expectation bias is also a thing.

In reply to John2:

I would be beyond amazed if this isn’t just a budget D-Link switch in a new case with a “respectable” brand attached to it.

 The clock won’t be doing anything - Ethernet isn’t a time-dependant protocol.  I would be amazed if it’s even soldered in. Jitter isn’t at all relevant because your streamed media is TCP, not UDP.

 I’m sorry you’ve been had, I really am, but this product is total snake oil.  If “EE” really stood behind their product, they’d provide measurements to support their spurious claims.

 For the cost (£500-£650!) you could actually buy some really nice network hardware.

In reply to m0ff:

In fact the 'English Electric 8' switch is a rebadged product made by the Chinese OEM 'ThunderData.' Exactly the same thing is licensed to other companies and appears under other brand and model names,  e.g. 'Silent Angel Bonn N8', 'NuPrime Omnia SW-8', etc. (Just google for all this.)

Of course, all these brands and model names are deliberately 'aspirational', in order to appeal directly to the target market.

As already discussed above, there is a chance that this switch produces less electrical noise than do comparable switches. But the markup being applied to the end price here is clearly a total swindle, in the same boat as the infamous Russ Andrews 'audiophile' power cables