HiFi snake oil or is it?

Blind test, or double blind if the switcher has any influence on you, like saying anything.

With coax digital interconnets it is hard to get the correct impedance at the plugs apparently – they should be BNC but aren’t. That can have some effect apparently if it is viewed as a transmission line.

But, whatever, blind testing is needed.

If you send your 0’s and 1’s from a compressed source , you need to decompress it.  If you repeatedly send the same song do you  train’ the decompression to do it better.? That is theoretically possible- don’t Spotify vary quality on demand theyre experiencing?

That you would, and do notice.  What’s the actual digital source?

You put some 0s and 1s down a wire, plus some extra 0s and 1s as a checksum. At the other end, you check and find that the checksum is wrong, so you ask for all the bits again.

Well, yes, Error Checking & Correction exists.  But the net result is the same, either the data is intact or it isn’t.  Reconstructing missing bits doesn’t change the outcome unless you’re having to use a lower bitrate to ensure a reliable connection.

SCART is just

SCART is a connector which can carry multiple different signals.  A more expensive cable might not just be better build quality but also carry discrete R/G/B instead of just composite video.

I think that’s more of a bandwidth issue in terms of steaming services like Spotify.

A bit like downloading a 200gb computer game… You can have the fastest internet connection in the world.. But the servers at steam for example, simply won’t allow you to download your game any faster than ‘speed x’ as they have thousands of concurrent customers downloading at any one time.

Ive never really tested it, but I’m sure they will also limit download bandwidth at source in times of heavy demand, such as when a new triple A game is released as they will have potentially millions of people trying to download it a the same time, so to avoid the servers simply crashing, they will limit the speed on a sliding scale depending on demand.

I will go against the grain and say “not snake oil” but I’m not going to claim that I can explain why.

Whoops.

What’s the actual digital source?

Amazon Music. I was using UHD tracks to compare 48bit/192kHz.

Indicate on the source and on the DAC.

I started this thread as a bit of fun, I was shocked at hearing an apparent difference between 2 bits of 50 cm wire.

There’s no need to start insulting people.

If you send your 0’s and 1’s from a compressed source , you need to decompress it.

Sure.  But that’s nothing to do with the cable.  Lossy compression is a thing, but the fact remains that what is sent down a digital cable must be the same when it comes out the other end or it’s unfit for purpose.

If you repeatedly send the same song do you  train’ the decompression to do it better.? That is theoretically possible- don’t Spotify vary quality on demand theyre experiencing?

Left to their own devices, streaming services will generally negotiate a reliable transmission rate.  A system is only as fast as its slowest component, even a regular old HDMI 1.0 cable will absolutely muller a TOSlink cable when it comes to throughput.  But would that cause issues with regular stereo audio?  I’d have to run some numbers but I’d have thought it highly unlikely, I’ve shoved uncompressed 5.1 PCM across an optical cable without issue.

Well, yes, Error Checking & Correction exists.  But the net result is the same, either the data is intact or it isn’t.  Reconstructing missing bits doesn’t change the outcome unless you’re having to use a lower bitrate to ensure a reliable connection.

Afaik that isn’t quite true for audio.  Yes there will be a bit of error correction but if it can’t be corrected then the player will interpolate between the two good points

https://www.soundonsound.com/sound-advice/all-about-digital-audio-part-3

After a quick google,

A PCM stream is in the order of kbps; optical digital craps out at 125Mbps; boggo HDMI is 5Gbps; the latest ‘ultra high speed’ HDMI is 48Gbps.

Whatever your audio issues, it’s nothing to do with digital interconnects unless it is broken.

For your testing you should use a local source, such as a CD, high bitrate MP3 or FLAC file already stored on your device.

You don’t want internet streaming/bandwidth to potentially add more variables/complictations to your experiment.

And just use one song, or maybe 3 songs that are very different,  say one Classical music track, one dance/techno, and one rock/metal track, so you get a fuller picture.

There is no bandwidth issues, both the streamer and the DAC tell me exactly what bit rate is going into them.

But the net result is the same, either the data is intact or it isn’t.

Cougar, old chap.

You’re talking about objective, measurable facts.

This isn’t a place for facts, this is a land for wires that are “subjectively better” and cost the price of a car.

With coax digital interconnets it is hard to get the correct impedance at the plugs apparently – they should be BNC but aren’t. That can have some effect apparently if it is viewed as a transmission line

Reflections on a 500mm digital wire will almost certainly not have any effect on the data integrity. Not at audio type rates. Might be an issue with 4k video speeds, but 48 or 96kHz (or whatever the current woo worshippers want in their system) – not a problem.

But yes, coaxes should be terminated.

Whilst a competent designer might put some internal termination inside the receiving unit, this is the land of audio woo and my guess (and my guess is subjective, so do one, audiophiles) you’re more likely to find an amethyst crystal embedded in the volume knob to absorb the bad vibes and negative karma.

Snake oil is actually an excellent conductor. But remember it won’t have as much value second hand, as it’s sound will have been affected by being burnt-in on two different systems.

I say keep the cable if it’s pretty. Think of the pleasure you could have tormenting friends by telling them it sounds better and asking if they can hear it too?

This isn’t a place for facts, this is a land for wires that are “subjectively better” and cost the price of a car.

:ROFL: emoji.

But remember it won’t have as much value second hand, as it’s sound will have been affected by being burnt-in on two different systems.

That can’t be right, the cable on the previous page comes with an optional burn-in service (for a premium).

Oh, dear god! They actually do. No emoji can adequately express my despair.

This isn’t a place for facts, this is a land for wires that are “subjectively better” and cost the price of a car.

I’m abusing that phrazing next I’m after a new ( unnecessary) guitar/ watch/ car/ bike/ Rapha item/ twinkly gadget ans have no actual need for it.

You’ve made my afternoon @mrmonkfinger

Similar but different…

My field is 230 / 400V electrics,  a colleague who looks after the IT side of things is currently doing his pieces that data cable runs in a newly finished building are too long. (10% less and it would be ok)

Being a sceptic, who has crunched on dozens of cat 5 connectors to cheap cabling to build a home network, i find out hard to believe that it matters (will Joe Schmo in the new building notice any difference when opening their word doc / tik tok video / etc).

Not saying it is not measurable, just not relevant in this case?

TLDR? Slippery (slithery) slope between audio and IT?

By the way fed up with slow Internet? Click [www.fishedin/sucker.com] to order an air purifier to speed up our home wifi.

@bigyellowmarin – The Cat 5 thing is real – all cables have a set of LRC characteristics; so we can use this to calculate the losses over a run of cable; for Cat 5 plugged into a normal port; it’s somewhere around 100m – and as your guy points out; 10% at that range makes a big difference. And that’s the thing about digital signalling; it works perfectly fine as the distance increases / the cable is poorer – until it suddenly stops. You might get a cable that just about works after error correction on the fringe; but chop a metre off and it works fine – add a metre on and it stops altogether.

But we’re talking cables that are two orders of magnitude longer here; and probably providing data at 1Gbps rather than 2-300kbps; again orders of magnitude.

For those that don’t already know; have a read about eye diagrams. https://incompliancemag.com/eye-diagram-part1/

Afaik that isn’t quite true for audio.  Yes there will be a bit of error correction but if it can’t be corrected then the player will interpolate between the two good points

That article is an interesting read, but it’s from over 25 years ago (and concludes what I’ve been saying all along, you won’t hear a difference until it breaks).

“The first thing to appreciate is the sheer amount of data created every second by a stereo A‑D converter.
…
If you scale it up for a 48kHz converter with 24‑bit resolution, an hour’s worth of recording would require just over 1Gb of storage. That is a lot of data to have to look after — and as yet we have not given any thought to adding error protection data or any auxiliary subcodes for copyright status or timing information.”

1GB might have been a “sheer amount of data” in 1998, but if my maths is on point a cheap TOSlink cable could pull that in about 2 minutes, ie you could retransmit it 30 times in an an hour.  Over HDMI you could transfer it six times over in one second.  Today I can stream 4k video in real time over the Internet for gods’ sake.

No way, no how is a digital interconnect cable costing four quid per centimetre making the blindest bit of difference on a modern system.  It’s a physical impossibility.  If what’s going in one end is what’s coming out of the other then there is no “better” to be had.

Not saying it is not measurable, just not relevant in this case?

This is kinda similar but kinda different.

Ethernet “Cats” are standards.  Cat5e – along with other Categories – specify things like maximum length of individual cable run.  If you stay within those limits then it is guaranteed to work, assuming everything is in spec.

Cat5e specifies 100m (including any patch cables at either end), if you have a 110m run will it work?  Probably, if it’s in an electrically quiet environment; or maybe not, perhaps it’ll only run at 100Mbps rather than 1Gbps, or start dropping packets all over the floor.  It’s difficult to say with any certainty.  How about 120m?  101m?  You’re out of spec, you’re on your own.

This is why I have a bee in my bonnet about Cat6.  The Cat6 standard is far more exacting than Cat5e.  If you’re doing a home install using Cat6 you’re almost certainly out of spec unless you’re actually a network engineer.  What you’re actually going to end up with is an install up to Cat5e standard only with awkward, expensive cable.  Which, y’know, is fine but rather pointless.

Coming back to AV: modern HDMI cables are certified.  If it’s got the hologram (and it’s not counterfeit) then a £5 Amazon Basics “Ultra High Speed” HDMI cable will perform exactly the same as a £500 one with the sheathing hand-rolled on the thighs of Filipino virgins.

just keep the cable, you know you want to

Well I like nice things, don’t care if I can’t tell a difference in sound. It was basically a freebie, I’d definitely keep it. Same if a paid thousands on speakers no way would I use cheap wires, everything would be in a compatible price range. Why ? Just because.

Just as easy to keep it and treat it like the free bag of haribo that wiggle used to send with an order, you’d be lucky to get £30 if you sell it as without full packaging and purchase receipt it could easily be an aliexpress knockoff

On your home hifi set up I doubt you could tell. This type is kit is usually paired with speakers and amps/etc that start in the 5 figures.

Don’t forget the wooden cable lifters!

Yes there will be a bit of error correction but if it can’t be corrected then the player will interpolate between the two good points

Ultimately all digital to analog reconstruction is interpolation, it’s just how smooth an interpolation do you want to do? FWIW I have a bit of experience with the display, storage, handling of seismic data, which ultimately is sound data, and how that is stored, and redisplayed.  I can immediately think of a good few things that are going to have a subtle effect on that reconstruction, and the quality of digital cable isn’t top. Ergo my loaded questions on how people like spotify handle the compression and decompression of data.

I’m not even sure if I’d trust blind testing – if you play a song once from a digital source once, then play it again with the same kit I don’t think it would be exactly the same as something outside of your control will have changed

There’s no point argueing about what should or should not be happening – double blind test it – if you can reliably tell a difference then that’s when you need to look for why there are differences, and if you can’t then that’s when you wonder abouot how powerful the various forms of expectation bias are.

I’m not even sure if I’d trust blind testing – if you play a song once from a digital source once, then play it again with the same kit I don’t think it would be exactly the same as something outside of your control will have changed

should really be double blind – unless you can find someone who is completely disinterested in what they are switching, or use some audible signal to indicate that it’s changed.

And you need to do the tests multple times with random switches to prove that you can reliably identify the better cable, or not. Don’t do things like open the windows/doors, drink a glass of water, blow your nose, etc in the middle of the test.

Thanks stevehine, thanks cougar, really appreciate your replies.

Would i be correct to takeaway from this:

If the install meets all the parameters of the Category the result is pretty much guaranteed.  Start abusing them and all bets are off.

Abuse could result in reduced performance (1Gb/s to 100kb/s to… ),

but the effect of the abuse is not linear, at some point the last extra meter will change it from a ‘far less than optimal but maybe ‘serviceable” to ‘plain broken’.

In a professional environment, it all comes down to risk appetite.

Not the risk of it not working, but the risk that lazy Shaun blames his lack of output on his spreadsheet taking too long to save, and everyone at his tribunal looking at you to answer “is it correct that his connection did not meet recognised  international standards”.

Dowotmate?

BYM > That’s broadly my understanding, yes.  I Am Not A Network Engineer.

As cable length increases, so does the likelihood of introducing problems.  Exactly what you’ll see and when I wouldn’t like to predict. Increased latency/lag, reduced speeds, retransmission, out-of-order or outright failure of packets… From the user’s perspective it’s going to be most noticeable if they’re streaming video or using Voice over IP (people generally have a low tolerance to even minor degradation of VoIP; a couple of extra seconds in saving a document would probably go unnoticed, a couple of seconds break in the middle of a sentence on a Teams call will rapidly have people climbing the walls).

A cheap nasty cable which is in spec but only barely might well see issues before a more expensive cable which comfortably exceeds it.  Some people argue in favour of full copper rather than copper-coated aluminium, the jury’s out here for me personally but if you are right on the edge of tolerance then it may make a difference I suppose?  If you’re running long cable runs parallel to mains power cables or through a machine shop full of lathes then environment could become a complicating factor (though really you should be running STP – ie, shielded cable – here).  And so on.

Conduction’s all in the surface layer Cougar.

Alu is almost as conductive anyway, IIRC.

Don’t forget the wooden cable lifters!

I knew using plastic cable lifters would come back to bite me.

Conduction’s all in the surface layer Cougar.

Aye, which is exactly why I said I’m unconvinced.

I can’t think of any other industry which routinely trots out such astonishingly abject bollocks and not only gets away with it but has a legion of fans who lap it up. 

Well, there’s this thing called Religion… have you heard of it?

Is it time for this again?  https://www.enjoythemusic.com/hificritic/vol5_no3/listening_to_storage.htm

I genuinely cannot tell if that is satire or an actual thing.

I genuinely cannot tell if that is satire or an actual thing.

“The various storage drives were sited in another room, connected to a Cisco Linksys E4200 wireless gigabit router in the listening room via 25m of Belkin Cat 6 Ethernet cable”.

25m of bog standard Belkin Cat6?

Connected to a Netgear GS108. If that doesn’t contribute to a dull and uninspiring sound, I don’t know what will.

Don’t turn valuable listening time into some kind of shite science experiment, just pour yourself a whiskey and enjoy the free upgrade!

I do also find that the more whiskey I drink, the better my equipment gets. Can any of the boring engineers on here going on about whatever boring physics they remember from 6th form explain that?? How does the whiskey manage to affect the signal from amp to speakers? I don’t know or care but perhaps it’s quantum.

I genuinely cannot tell if that is satire or an actual thing.

That has to be satire otherwise nothing makes sense any more

No problem with the OP raising this.

This is my take on this particular situation.

Some where in the uk a file of the song i want to listen sat on a server. It’s probably 100s of miles away. When i listen to that song you send a signal and that server starts sending the song to you. The 1s and 0s come off the hard disc as electrical signals. These are then converted to flashing light and go into a fibre optic. At least every 30km the signal goes back to being electrical to be cleaned up and boosted. It will also be converted to electrical signals for rerouting. Terrifyingly all the songs 1s and Os are multiplexed (mixed up with) everyone else’s 1s amd 0s in each fibre. Eventually it gets to a green box in my road and is then converted to an electrical signal for the last time. It then goes into a 1950s pair of copper wires. This wire then goes down my road under ground then up a telegraph pole, across the road to a horrible leaky plastic box hanging off my gutter. It then comes in though the front door to another horrible plastic box. It’s then about 3m to the router. Now my mate knows about this stuff and says the industry standard is that with error correction I’m getting only 1 bit wrong in every million million bits (10^12). Now you might be able to interpret between missing bits in songs but not software. I can save have downloaded things like photoshop and and they work! So this bit I’m struggling with. Having made this incredible journey the bit where the digital signal gets lost and degraded and we have to start interpolating is the last 50 cm.

You can see why I’m sceptical

Despite what your kit is telling you Amazon music might be a source of variation . They’re sending you a compressed signal – do they guarantee a rate? I used to be able to tell the difference between good and bad days on Spotify,  but going to premium fixed that as you go from streamed ( variable quality) to a download

Well I think I’ve come to my own conclusion.

I bought a 3rd cable for £12.

So I have a £5 cable, which I originally ordered to replace the £200 snake oil cable, as I did not believe it would sound any different.

I still think I can hear a difference in the sound from the cheapest cable but I can’t from the other two.

The £12 and £200 cable sound the same.

So there seems to be something going on with the cheap cable. It can go back to Amazon and I’ll get my fiver back.

I’m still not 100% on any of this and do wonder on the placebo effect of it all?

I didn’t mention in my OP, that it also came with a fancy 3rd party power supply, still retails at £300.

I replaced it with a £35 Audiolab one. the thing is, the stock one by all accounts, a very good quality unit and supplies both AC and DC for different parts of the DAC.

It makes no difference.

There is even a fancy, super chunky, £50 mains cable.

So I’ll be punting the 3 on ebay, a value of £550 new.

Even if I can only get 40% of their new price, (on things in good condition it’s normally 50-60% IME), it will greatly decrease my original outlay of £260 for the initial purchase.

I wonder if I have just missed falling down the seperate ring main, mains conditioners, cable lifters, rabbit hole?