How much does it cost to be an audiophile?

[mansr]

31 minutes ago, gmgraves said:

I think a perfect example of that are the MSB Diamond series DACs. They are incredibly expensive and they use ladder DACs (also called R2R, and Multibit DACs) of their own design and executed in discrete components. Then the whole megellah is potted in epoxy and housed in a gold-plated aluminum "brick". I read in an online forum, recently where somebody tested an MSB Diamond V DAC recently against a Benchmark 3 with the new Analog Devices SabreDac Pro, and found that it both measured and sounded better. It seems that all of that "unusual circuitry" buys one (at least in this case) is an astronomical selling price! with today's semiconductor technology, one shouldn't assume that discrete circuitry is better than ICs. It was once true, for sure, but you can't assume that it still is anything other than more expensive. 

ICs have a huge advantage in component matching as well as thermal coupling, both very important for an R2R DAC. There is no way a discrete design will get even close to an IC in this application.

[mansr]

1 hour ago, sandyk said:

Very interesting.

The video wall was a fun project.

[mansr]

 

This image shows the built-in AFG of my Tektronix scope outputting a 10 MHz square wave through 0.5 m RG-58U to the channel 1 input, then via T coupling and 2 m RG-58U to the channel 2 input which is set to 50 Ω termination. The somewhat distorted top of the yellow (channel 1) trace is due a slight impedance mismatch somewhere causing reflections. Other than that, the only discernible effect of the 2-metre length is a 10 ns delay. I think this is good enough for audio.

[mansr]

12 minutes ago, sandyk said:

It's not about losses due to the cables, it's the effect of the additional capacitance seen by the output device, and many opamps do not like to see a lot of direct capacitance at their output and may even become unstable, which is why most output I.C.s in typical consumer gear   (e.g.LM4562, LME49720 etc.) have something like a 100 ohm series resistor at their output.

If the added capacitance is that troublesome, then a longer cable would be equally damaging. I don't recall seeing a maximum interconnect length specified in any equipment manuals.

[mansr]

1 minute ago, sandyk said:

The series output resistors take care of that problem.

Then what's the problem with the suggested setup?

[mansr]

5 minutes ago, semente said:

More components in the signal path?

Only a slightly increased capacitance, equivalent to a longer interconnect.

[mansr]

2 minutes ago, sandyk said:

Mansr

 Something specific.

Some constructors of the attached reported instability problems, and even overheating of the device due to oscillation.

 This was apparently caused by the direct capacitance of the lead to the RC volume control. One cure was to replace the output capacitor with a 100 ohms resistor. Due to the low DC offset this didn't cause any rustling noises etc. with rotation  of the control.

Alex

I'm not disputing this. What I'm saying is that if a component can drive 2-metre interconnect, it can drive two 1-metre lengths in parallel.

[mansr]

1 minute ago, sandyk said:

Obviously.  But the interconnects used as George is suggesting may need to be longer than 1M.

Why?

 

1 minute ago, sandyk said:

From what I see in this forum, many use interconnects considerably longer than 1M due to equipment layout considerations.

Yes, I've seen Mani's 10-metre runs. So what? Don't do the test with extreme lengths, and it will be fine.

[mansr]

Just now, sandyk said:

Mansr

This wasn't meant to be an ACADEMIC test. It was to enable your typical user to choose between 2 different interconnects, which in many cases will be much longer as you just pointed out. This will likely affect many typically implemented opamp output stages to some extent, and neither cable will sound quite as good as it should.

If you normally use extremely long interconnects, temporarily move your equipment for the test session. I doubt many would be willing to drill extra holes through walls and floors just for a test anyway.

[mansr]

13 minutes ago, sandyk said:

 You are missing the point.

The object of George's proposal was to enable selection between 2 cables to use with their equipment at it's present location.

 The cables they will wish to choose from, perhaps with a view to purchasing, will need to be of the length they normally need to use.

Well, most people don't have insanely long interconnects. Maybe George's suggestion is unsuitable for a select few. It's still a good method for most. You're trying to invalidate the entire idea on the grounds that there might be a setup where it doesn't work. That's not sound reasoning.

[mansr]

3 hours ago, esldude said:

Also I think we are mixing up engineering approaches with science.  There are some subtle differences. When enough is known dropping the last little bit of scientific rigor for engineering practicality is very reasonable and efficient.  The working of audio signal over cables certainly falls in this range. So much so any who disagree have the burden of proof.

Engineering is the art of knowing what to ignore.

[mansr]

7 minutes ago, Jud said:

But not the science?  

Science gives us the tools to calculate everything in minute detail. The trouble is that for real designs, such a detailed analysis is unpractical. Engineering is all about simplifying the science to make it practical while still resulting in a working product. For example, a complicated calculation can often be replaced with a much simpler approximation guaranteed to be greater (or smaller) than the exact value. Suppose the task at hand is to choose a wire capable of carrying the current required by some apparatus. Calculating the exact current might be hideously complicated, so instead we use an approximate value that is at least as large as the real one and pick a wire accordingly. If the wire gauge we arrive at is still practical from a mechanical (and cost) perspective, we're all good. It doesn't matter that a slightly smaller wire could have worked.

 

7 minutes ago, Jud said:

A friend who won an award a few years ago for best civilian engineering project for the Department of Defense has many stories of fellow engineers who felt various factors could be safely ignored in their work, and were proved wrong when the end product failed.

A famous case of ignoring the wrong thing is, of course, the Tacoma Narrows bridge collapse of 1940:

 

Modern bridges include damping elements to prevent such uncontrolled oscillations.

 

7 minutes ago, Jud said:

Of course those make the best stories; I'm sure there are also  plenty of examples of people obsessing over things that ultimately don't matter.

Yes, this forum is full of them.

[mansr]

27 minutes ago, gmgraves said:

In fact all you can find on the subject are "white papers" written by companies whose self-interest is tied up in this controversial subject being real. Unfortunately, if you read these self-serving "explanations" with an educated eye, it's clear that what these white papers are saying is either sheer double-talk or is alluding to real-world phenomena that only affect coaxial cables in the VHF or UHF regions, not in the audio spectrum.

In addition to this, many of the claims put forth are in direct contradiction to one another. Until they can at least get their story straight, I see no reason to pay any attention whatsoever to them. Other than for entertainment. obviously.

[mansr]

3 minutes ago, sandyk said:

How many E.E.s etc. completely reject the possibility that a gadget such as a USB Regen is capable of improving USB Audio

Would it surprise you to learn that I have recently been sketching out a simple USB widget that could improve things in certain situations? No, not the power monitor I already posted about.

[mansr]

1 hour ago, pkane2001 said:

Mani's ABX test was a good example where there was a bit of confusion on the goal of the test, the test set up, and test procedures. Nevertheless it produced some interesting results, which only a few of us actually spent some time trying to analyze.

 

While I didn't find a smoking gun, so to speak, I did find differences in the audible range between analog captures where there was a software configuration difference and no such difference where there wasn't. Maybe that explains the ABX results. Most likely it's not new science being discovered here, but perhaps some new knowledge about Peter's XXHighEnd software and the SFS setting.

There is also this possibility:

[mansr]

12 minutes ago, pkane2001 said:

Good point on training, Jud. I suspect there's a lot of room for better tools and materials for those who are curious. Know of any that can simulate various distortions, jitter, etc?

Matlab.

[mansr]

38 minutes ago, semente said:

Some years ago, after hearing a slight difference when reversing the polarity of the mains cable (easy job with Schuko plugs)

There's nothing mysterious about that. There's always some amount of capacitive coupling between the chassis and the mains input. When components are connected together, small ground currents will form. These can be amplified into audible hum, especially by high-gain phono preamps. This is one reason for using balanced interconnects with the shield connected to one end only.

 

38 minutes ago, semente said:

I started using a VdH gizmo to set polarity.

 

https://www.vandenhul.com/product/the-polarity-checker/

Overpriced.

[mansr]

20 minutes ago, davide256 said:

That a very 1800's or DC power view of electrical conduction, kind of ignores quantum dynamics. When dealing with AC current, skin effect applies for conductors and junction effects for the joining of two different metals ( cable and terminating cable plug)

Maxwell and Heaviside did their work, including research on the skin effect, in the 1800s.

[mansr]

1 minute ago, Jud said:

So it's one of the things we miss out on when we dismiss accounts of changes in sound due to cables

Nobody ever said changes in wiring topology can't alter the sound.

[mansr]

58 minutes ago, semente said:

Did you see the news?

 

A Frenchman has won a million euros twice in the last eighteen months on the same lottery, media reported Wednesday, a feat mathematicians said carried odds of around 16 trillion to one.

Be careful with such statistics. The odds of somebody winning a million once are pretty high, depending on the format of lottery possibly as high as 100%. The odds of a winner winning again are exactly the same as for any given player to win once. The odds of the same player winning in two specific draws are indeed extremely slim. The odds of winning twice in a larger number of draws are considerably better.

 

Still, he's probably the luckiest man in France. That is, assuming being in France can be considered lucky at all. ?

[mansr]

1 minute ago, Ralf11 said:

what's wrong with France?  ?

It's on the wrong side of the English Channel.

[mansr]

2 minutes ago, Ralf11 said:

careful  -   the Normans may invade again (not for the food or culture, obviously)

Again? They never left.

[mansr]

6 minutes ago, semente said:

The food is rubbish here, a gastro-invasion would be most welcome.

You must be eating at the wrong places.

[mansr]

 

[mansr]

8 minutes ago, firedog said:

Every time I visit there I eat fine food. Lots of fresh locally sourced products. I guess it depends where you eat. There certainly is a lot of poor food there, if you want to buy that. 

I have never been to a place where you couldn't find bad food.