Sound Quality Vs. Measurements

Status
Not open for further replies.
Unlike Cheever, Hawksford did some excellent work as well. He's an enigma: I always find it difficult to understand how the same person can produce fine analyses and then turn around and crank out junk without blinking. But "fuzzy distortion" has a nice memetic slogan feel to it, so it's useful for browbeating purposes.:D

Barrie Gilbert doesn't get it either, nor do I. The wire and fuzzy distortion stuff make no sense at all.
 
diyAudio Member RIP
Joined 2005
I know a lot of PhDs. One thing I can say about them - they all have degrees!

Me too. I have worked with them, drunk with them, and even slept with them :eek:

And spending only, typically, a few additional years in school to have the incredible distortions of self-regard so often induced by the conference of that title is perhaps a bargain of sorts.

But is the result worth it? For a few who are truly brilliant and creative and highly motivated to begin with, she or he will probably do what they do anyway. Although, admittedly, funding may be more difficult to obtain.
 
About Mr K of B&K. While I am impressed by his ability to resolve low level distortions, I am deeply suspicious of the correlations he seems to be making. His blind test results are meaningless if based on the wrong premise, that is, low level distortions or whatever else is giving rise to preference. Since the tests are conducted between two different amplifiers he has immediately invalidated the test. The test can only be valid if the only thing that changes is the low level artifacts he says are responsible for preference.

An interesting opportunity arose to compare a number of different amplifier topologies. A valve amp manufacturer had completed a design of a SET which one of his dealers had asked for. The normal valve products from this company were high power low distortion monoblocks. I had at the same time, in the same building, completed a prototype 450W MOSFET monoblock. So we set them up in the listening room which had B&W 801s wired with silver cable and a Meridian CD player as source.
First up was the 8W SET. Initially it sounded very rich and pleasant but after a minute or two it became apparent that significant detail was being masked, and your attention was drawn to it.
Next were the valve ultralinear 100W 0.1% THD monoblocks. These sounded clean and the improvement in detail over the SET was marked.
Third were the 450W fully balanced MOSFET monoblocks. These marginally had the most detailed sound but were dry in comparison with the valve monoblocks, but had better bass. The story could have ended there but I decided to unbridge a monoblock and run it as a stereo amp. The circuitry was unchanged, I just rewired the input and output to create a 150W stereo amp. This amp produced the 'best' sound of all. Meaning, it was the most engaging to listen to. This amp had a valve input stage and the design was crafted to allow pure 2nd harmonic which would cancel in the fully balanced bridge mode of operation. In the unbalanced stereo mode it was giving 0.15% THD. With this amp however, I could just detect loss of detail.

The last part of this casual test is the type of test needed to test a premise about quality perceptions. Ie. in this case the electronics do not change, only the mode of connection giving rise to a different set of distortion conditions. Ideally the connection should not change only the distortion, but this is as close as practicable.
 
seems to me John you dismissed a peer reviewed article specifically addressing the subject and launched an implied insult to the integrity of a PhD just a page or so ago

Well, PhDs, no matter what their field or even profession is, do share some common traits.

Many are teaching somewhere, and they cannot escape the effect of looking at young heads absorbing their words for 30+ years, making them start to believe that they are more than human. :D

When in trouble with argumentation, they tend to use the Ultimate Weapon - but I have a PhD! :D

If asked to do something practical, rather then theorize, they take their time about it - I suspect they need it to resolve all the conflicting requirements which come to their mind.

I am always reminded of the immortal scene in "Ghost busters", when Dan and Bill are sitting on the staris, after just being fired, and discussing their options. Bill says: We could try the private sector. Dan, much disgusted responds: No, not them, they want results!

With all due respect to everyone here, just look at the last 5 pages - I wonder how long does it take some people here to resolve all their internal conflicts in theory before they actually solder a transistor to a resistor, for fear of what that might do or cause, discovered or not.

Why not invert this, and FIRST make what you need to make, and THEN start cleaning it up from unwanted effects? That seems much more sensible to me, unless you are actually breaking new ground, when it becomes much more complicated, because then you become Otala, and the wolves all come for their pound of flesh. :p

As things stand, turning anything mentioned into a theoretical clash of the titans, won't get much done in vivo.
 
Well, I am still interested, dvv, in your amplifier model. Best not to argue. '-)

So am I, John. :D

Did you pick up the final form version some pages back?

My PCB artwork buddy is currently up to his neck in work, so I'll have to wait a bit, but it should be done by the end of May. I'm rasonably good at it, but he's MUCH better, as you'd expect from a guy doing it professionally for the last 25 years. I hate meddling in anything when I have a better man for the job at hand, and in the immortal words of Clint Eastwood, "A man's got to know his limitations".

Actually, as you said yourself, I never manage to get it all just right as per Otala. Part of the reason for that is my extremely limited parts selection. For example, Thorsten told me even he has problems finding the Hitachi lateral FETs he used for drivers, and that's saying something, Germany's market is extremely well supplied in general.

At this time, I'm not very happy with an open loop bandwidth of 63 kHz. Not that it's a poor bandwidth, it's more that I have a nagging feeling I could, and therefore should, have done better still. Can't escape that feeling, so I'm using the extra time to investigate.

I'm probably overdoing it, I tend to be like that sometimes, but I can't help it. If I am not proprely driven, what's the point, I might as well just buy an amp.
 
Speakers .

David Mate who I think I have referred to told me that in his opinion one of the big problems is that humans are too good at coping with phase errors in speakers . Two speakers that would be laughed at if amplifiers would be said to be similar because square-waves from speakers is not very important . David suggested that processing the information was fatiguing . He did say crossover at a minimum of 7kHz if making a speaker accepting phase is difficult and budget limited . Another whose opinions I use as reference points said if listening fatigue was real then visual fatigue would occur as the bandwidth is greater with vision .

Loosely connected is this . I was asked recently if Jean Hiraga had written a mathematical paper as by the way I was speaking he had . I said yes in a way . Hiraga said he thought hi fi that sounded pleasing would have harmonic distortion which reduced on either a linear or exponential curve . He had no idea which . I always remember Hiraga gave Coral 777ex as a pick up that fitted this rule . I used a 777 ex at the time so studied Fourier series after that . I have always asked myself which curve fits it best ? I was in a pub one day when a man got very angry with me after he overheard my conversion about this with a friend . He insisted that we knew the best curve that described the stock market . I tried to say about Chaos theory . No good , Chaos theory had been dismissed on a forum he was a member of . In desperation I said I did know actually . It is the Fibonacci sequence . He was delighted and went home . I then said to myself , is it in audio ? I am so glad I am not on a stock market forum . Sorry about punctuation . My hero Alan Blumlein apparently struggled with it .

http://magneticmusic.ws/NFRhufker.pdf
 
Last edited:
@wayne

I read the interview with Bob Cordell, Wayne.

Not surprisingly, he started out with a poke at Otala - the man is obssessed.

But for the most of it, I would agree with him in general, most of the differences really being only shades of gray, rather than essential points.

I especially agree with his views on what he calls "boutique power cables". About a year ago, I decided that talk had done all it could have done and that it was time to check it all out in hard practice. So I went visiting to a friend's store, the biggest source of high quality cables localla. He stocks literally like 30+ tyoes of each connecting and power cable known to man.

However, the core of his business are Neotech's cables. They are a company from Taiwan, manufacturing cables and wiring of all types. You may not have heard about them or their products as such, but they are the actual cables behind quite a few VERY famous names, with even more famous price tags. As an example, they will do you a cable of your choice for $10k, write whatever you like, pick a color from their palette of colors. Britain's Ecosse cables do just that, and their Big Red power cable, length 1.5 m (app. 5 ft), with Furutech plugs on both sides, will set you back around $800 approximately. The same cable, also with same Furutech plugs, not red but black, and without the "Ecosse" brand name, cost me less than $250, also in full retail.

How do I know they are excatly the same cable? Simple - we (a few of us were very interested) took a piece of the Ecosse cable, and a same lenght of original Neotech cable, and both measured and auditioned them, under the same conditions, same room, same systems, etc, and all we got is twice the same thing at two darn different prices.

I would be less than honest if I said we got the same results with either one of those cables when compared with standard, thick standard copper only power cabling, available just about everywhere, for small money. A part of that was due to the Furutech plugs, which are engineered much better and to closer tolerances than the usual fare - we eventually produced a standard cable terminated with the same Furutech plugs. They simply make for a better contact, period - whether their price justifies the end difference is a matter for debate, and I think is best left to everyone to decide for themselves. Too much placebo in that subject.

And anyway, given that my dad was a mechanical engineer, I am prone to as best mechanical quality as I can get, which is not always rationally justified, so I'm not the one to suggest anything on the matter.

After having built my own car literally from the ground up, I am hardly the one, and I was moved to do that beacuse I wanted a tailor made vehicle. It's an Otala design, it slews from 0-60 mph in 6.5 seconds with only a 1.6 litre, naturally aspirated engine. :D AND, as a free bonus, it looks just like a Yugo, which especially infuriates my customers. :D:D:D Priceless! :p

The rest of the audible difference in comparison with standard thick cables is so small that I seriously wonder if it's imagined rather then real. Whether it was placebo effect or not. But the very fact that I cannot determine that sort of kills the boutique cable reasoning dead.

Thus, I agree with Cordell on that point especially.
 
Last edited:
Status
Not open for further replies.