Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[iain]

This is a quick question for Dan lavry or anyone else with an opinion, I know you've all been asked this type of question if not this question every day for 10 years, so ill write every thing I think I know, and hope fully give you a yes no answer to make.

I study acoustics (results based on science tests and figures), and work in a hifi shop (results based on listening and opinion), and so always find myself stuck between a rock and a hard place when discussing the merit of audio equipment.

I am currently debating with a work college whether a

[Barry Hufker]

Iain,

You'll get better postings from people other than myself but I do want to jump in.

24 bits only equals a dynamic range of 144dB in theory and not in practice -- at least not at this stage of digital reality.  I am confident the DAC itself and the analog components have much more of an effect on sound quality than an expensive wire.

From a testing standpoint you're at a stalemate.  He says the wire makes the sound better.  You question that.  There has to be at least a third party to join one side or the other.  But more importantly, could your friend tell the difference if he didn't know whether the wire is in the circuit or not.

As long as the digital information travels the length of the wire intact, there is no way to improve the digital information no matter what wire is used.

If I paid 400 GBP for a wire I'd sure as hell tell you I heard a big difference whether I did or not.  No one wants to appear a fool.

Barry

[Scrith]

Can a blind test be arranged?  Many transports include at least two outputs (coax and toslink, for example), and most DACs have at least two inputs with a switch in front to pick which one is in use, making comparisons really convenient.

I've done this a number of times to compare coax vs. toslink in my setup.  At one point I think I had a really bad toslink cable (one of the eBay glass toslink cables that many people claim is a bargain) and I could actually tell the difference between them (with my wife switching the input randomly on my DAC).  But after purchasing a reasonably priced replacement (Van den Hul Optocoupler) I can't tell the difference between them anymore.  I have some suspicions that something can happen to that glass fiber (micro-fractures or something?) that leads me to distrust toslink, but I'm fairly certain it would be difficult to hear the difference between coax cables.

By the way, I'm generally a skeptic of blind tests, based on my own experience.  I've done some careful tests on my own to try to tell the difference between two cables, only to have them end up sounding the same.  So then I start using a new cable...and a few days later I hear something different in some random piece of music.  I then pick up the blind-testing again using that music, and can sometimes hear the difference, at last.  Is a slight difference I can only hear every few days worth an expensive cable?  I'm not sure...  I also wonder if there is some sort of subliminal audio information in there that isn't so obvious (and can't be detected reliably in a blind test), but actually makes you feel differently about music over a long listening period.  I'm sure somebody has something a lot more detailed to say about this type of thing.  I'm also sure there are cable skeptics out there who confidently dismiss such ideas.  I maintain an agnostic position on cables for now.

[Gunnar Hellquist]

Iain,
    I think you either has to believe in the difference or change job.

Personally, I am firmly convinced that good enough is what it takes for digital signals. Any more than good enough is not adding anything from an engineering standpoint. Less than good enough and you most often get very recognizable errors. There can be quite a number of different criteria for this -- on an optical interconnect signal damping is of course important. Mechanical stability, resiliance to fire or oils might be other factors, mostly not relevant to a typical home setup. Personally I use the

[iain]

Hi Barry, he as yet hasn’t paid for it, thats the liberty we have as hifi shop employees we test stuff to death before we buy it, but after testing to death he’s still considering buying it (although at trade not quite as bad).      

With regards to Scrith’s comment when I blind test it, and test able to see the cable I can’t hear what I would describe as a marginal difference, and I’ve tested my ears in anechoic conditions I was able to blind test identify a level change of  1.5 db + my hearing still goes up to 16.5 kHz.  I shall continue to listen with more DAC’s, more amps, & more CD players and see if I can hear anything ill let you know if I do.

With regards to Gunnar im defiantly in the believer camp I believe that music reproduction is still a long way from being truly realistic even on 10k systems, But I don’t think throwing money randomly at parts of the system is going to help. I think we need more standardization as I say, is there not enough members of the AES on the web board alone to forward some kind of motion.

i.e.

Measurements are in two categories – AES test and Manufacturer test

For test can we not concoct a particular test and set of  measurements that would adequately demonstrate the performance of a DAC for example, we certainly have measurement mics at my uni that, shock horror are more sensitive than the human ear. I’m no expert in electronics but I imagine it is possible to measure changes of 0.0000001volts the voltage equivalent of the quietest thing we can hear. People may then say well a sine tone and pink noise are not music, they have no transients, no dynamic range, so create a peace of music that has the maximum possible dynamic range, frequency content ect and test and analyse that.  There’s, signal in, and signal out, any changes are a result of the DAC, im sure the modern FFT & distortion analysis is up to the task. From this derive say 4 AES measurements i.e.
Averaged A-weighted THD
Averaged A-weighted Signal to Noise
Averaged A-weighted Channel separation
Averaged frequency response variation

(or what ever else the AES members feel is necessary)

(I say A weighted as the A weighting curve is a reasonable middle ground and has been shown in the acoustics field to best represent peoples average perception of loudness and hence optimising to this should yield the best apparent audio reproduction I think .)

So every ones test spec pages look something like this
AES measurements

Averaged A-weighted THD
Averaged A-weighted Signal to Noise
Averaged A-weighted Channel separation
Averaged frequency response variation


Then the AES provides clear guide lines on the test procedure, and these figures should be able to be replicated by testing the relevant peace off kit by any company or lab that has the facilities, If your competitor is making suspect clams about his kit, test it and make sure every one knows there false. 20 companies come forward and say those figures are wrong, im sure the public will know who to believe.      

+ specify AES Cable resistances lengths ect (which hopefully everyone’s agreed on optimum)
If your kit needs special cable specify it, or provide it, don’t let the consumer be coned by other companies, who frankly haven’t earned his money.

Then 'Manufacturer test' can contain whatever condition the manufacturer likes i.e.  

ZHR weighted THD + signal reduction factor@1db gain of 63 = 0.000000000000001dB

With test standardization would we not drive manufacturers to produce better equipment not better marketing.

As to my job I would like kit that I can say and know will give the customer a good and functional product, yes someone may be happy with the idea his 10k system has improved but he’d feel a lot better and so would I if it actually did.

Thanks for all the feed back everyone, I think ill rephrase my question to Dan

What would be the Approx manufacturing cost of a 2 meter digital cable that you feel would allow the DA 924 to operate at its optimum capacity.

p.s i took this picture in scotland at the weekend



Cheers
Iain

[Barry Hufker]

Toslink is a different question.  That's pretty much junk.  The telephone company does that technology correctly -- very high grade optical glass and high quality transmitter/receivers.  Toslink "optical fiber" tends to be extremely cheap plastic with a lot of "loss" -- poor light transmission down the fiber.  I would expect real problems there before any system with a good quality wire.

http://www.gotham.ch


GAC-1 S/PDIF-Pro (10070) Silvercable -- 1.85 Euros per meter or 146.5 Euros per 100 meter spool.

Barry

[blue2blue]

There are some folks  who seem to think that digital audio systems are always scudding along the brink of audio disaster, beset by torrents of uncorrected transmission errors that continually degrade the signal  resulting in all kinds of psychoascoustic nastiness.

I just don't think that's the case.

The job we ask of the digital part of the digital audio equation (in between AD and DA, as it were) is pretty straightforward and is not exactly a novel technical art.

It's easy to test transmission accuracy and that's how we've determined our standards for various interconnects. Those standards are not based around outside performance limits or best-of performances but rather are developed extremely conservatively to assure maximum accuracy.

There are so many places where we have to accept ambiguity and even mystery in audio (at least in the short term) but I just don't think the performance characteristics of digital cabling is one of those places.

[Gunnar Hellquist]

Hi again Iain.

Seems like you are willing to talk to the sceptictal people. Sorry if I destroy some beliefs. My suggestion though is to not use this kind of talk with your collegue, it will all only end up as a mess.

Your original question was something like this:
> How can a

[trevord]

Just to chime in..
I go with - no damn difference!! (to be subtle)
Gunnar says it best.
A point on jitter.
Jitter is important in ADC and low latency DAC.
Low latency DAC is used mainly by professionals where you turn a knob and want to hear a difference immediately.
Most consumer digital applications have enough buffering before/after cable transfer so that digital cable induced jitter (unless its so bad it destroys the data) does not have much of an effect.
Of course there is a lower limit to the type of consumer app I am talking about.
Who knows what the really cheap crap is doing

Once a cable meets specs, you can't get better.

There may be a justification for cost tho.
A typical cheap digital cable is built from spec'ed components which have variation in quality.
A low cost manufacturer will assemble and assume since it was built from spec'ed parts - then it is in spec. Of course cheap materials for cheap cables will mean they have a variance (lets say 20% for each component) which means the final product will have a lot of quality variation.
Maybe no testing is done after the cable is built.

A reasonably priced cable will be built from materials with 10%  variance in parameters.
A sample of cables is taken from every batch to test to make sure they are in spec.

The most expensive cables (justifiably) will be built from materials with 5% or less (number are relative only - don't quote)
BUT
if the manufacturer tests each and every cable - the cost would be astronomical (relative to the cheapo).

The same may apply to the same brand - that is a manufacturer may  build with cheap materials and test 50% of them. Those which meet spec are the high priced versions , the others are put back in the "bin" with the rest.

So there is some variation in price. But not because a cable is "better" - a price increase may only mean - every cable of a particular brand meets spec. A lower cost may mean 40% or more of a particular cable is out of spec and you take your chances.

I am not saying this is a bad thing. You get what you pay for.
But you can't get better than the specification.

[Barry Hufker]

Let me just qualify one of the last statements -- you may get what you pay for.

Barry

[ssltech]

A wise man once answered a similar question in the following way; repeated here with the (paraphrased) question first:

Q] -Surely it's not possible to hear the difference between a $400 digital interconnect at my local hi-fi store and a $10 digital interconnect at the same hi-fi store?
A] -You CAN hear a difference, but only if you stand the side of the counter where the cash drawer opens towards you.

Depends what you're comparing it to. I personally reckon I can detect the difference between an expensive interconnect and a short length of twisted wet string, but as soon as the wet string is replaced by something which passes the data in a manner which allows the words to be reconstructed reliably (which costs all of about $5...) then I cease to be able to hear the difference.

Keith

[maikol]

I think trevor has a point here in the "why can it be so expensive?"


I once read all i could on "Transparent cable" 's (could be the maker of that 400GBP digital cable BTW?) website, and on Shunyata's too, just trying to see if their was some real technical info.

I have not found any, of course, but i saw that their most expensive cables each need several hours to be done (at least it's what they say!).

If that is true, then here is a big part of the answer considering the price.

I would compare that to buying a watch. You can buy a cheap quartz chinese watch for a few bucks, or you can buy a Patek Philippe, Breguet or whatever for several thousands. Both will give you what a watch is made for. The several thousands one will never be more precise than the cheap one (if not less because it might be mechanical!), but if you buy it, you are probably looking for more than just accurate time info...

And if you have such a beauty at your arm, you'll certainly have to have an expensive car too, and an expensive hifi system, and... you name it...

This is just a part of the luxury business!

And oh! BTW i'm one of those who believe there are differences


well maybe not between digital cables!

Michael

[iain]

Again Thank you all for the feed back

An update on the cables my fellow staff member asked the area manager (next in line to him is the managing director owner of the company) today why digital cables made a difference and then made me explain what I was saying to him (the area manager), that was kind of awkward, but anyway what I eventually got him to admit (although I probably get a wage cut now) was that a modern well designed dac should not be improved by expensive digital cable, we then had a test with a (im going to mention names but only in positive light this in my opinion is all very good equipment and worth every penny)  2k musical fidelity dac, + a 3k denson transport + a 3k musical fidelity pre and power + 2k dynaudio speakers, 10k all in we’ve got more expensive stuff but we like the sound of  this lot. We tried digital coax from

[Barry Hufker]

If I answer correctly to any of the above do I get a 400 GBP interconnect?

Barry

[Gunnar Hellquist]

Hi Iain,
sorry about the wage cut. But I did try to warn you.

Gunnar