Sound Quality Vs. Measurements

Status
Not open for further replies.
One 'resume':
 

Attachments

  • Picture 23.jpg
    Picture 23.jpg
    551.8 KB · Views: 117
Well, they were both professors in the department of psychology at WWU for a long time, and they both had interests in perception. Do that need the 'seal of approval' of Lipshitz et al to be taken seriously?

Perception is a different area of study than sensory science. Which is one reason why this fluff piece had exactly zero impact, other than to give you another 30 or 40 year old piece of paper to wave around.

just 'lurks' these days, so far as I know.

And teaches Ayurvedic something or other in conjunction with the gin rummy course.
 
AX tech editor
Joined 2002
Paid Member
Well, they were both professors in the department of psychology at WWU for a long time, and they both had interests in perception. Do that need the 'seal of approval' of Lipshitz et al to be taken seriously?

If they knew so much about perception, why then did they limit themselves to what is 3 pages of: 'some say this, some say that, they all have a point, hopefully we can put it to rest'?

John why not READ THE F PAPER rather than just the CVs? You know in detail the CVs but you miss completely the contents of the paper.

Jan
 
Haha, you crack me up! That's just bar talk, arguments we've heard over and over again, and they skilfully skirt the issue of psycho-acoustics, and why a controlled test can take care of all of those factors.
Though I chuckled when they called on Gordon Holt, the man who went on record saying:

"Audio as a hobby is dying, largely by its own hand. As far as the real world is concerned, high-end audio lost its credibility during the 1980s, when it flatly refused to submit to the kind of basic honesty controls (double-blind testing, for example) that had legitimized every other serious scientific endeavor since Pascal. [This refusal] is a source of endless derisive amusement among rational people and of perpetual embarrassment for me".

Another thing that made me smile was the attribution to a statement by you, that there may be personal reasons to hear what's not there, or not to hear what's there. You forgot all about that?

Jan

Yes, Gordon did say that. However, he did not do such tests himself while evaluating equipment. He had a keen ear and aural memory, and did extensive recording of live music. Gordon was a true expert in hearing flaws in components, and was totally honest.
 
rare skills if true - how to establish their factual basis without testing with controls? - lots of anecdotes out there - to support just about any position

those who rankle at the idea of "self deception" being involved in expectation effects are denying basic human conscious perception - its not a choice, not about "honesty"

it is about how brains have evolved, how perceptual input is filtered/massaged/warped even suppressed way below the level of conscious awareness
 
One way of approaching the listening and differentiation 'problem' is to learn to listen to, and for distortion - most audio systems fairly stink of it, :p, so pretty easy to find! An easy one to start with are the sound of cymbals - which many systems get hopelessly wrong; in fact this is how I started on my current audio journey, :).

Take a very average older rock recording, and just listen the quality of the drummer's cymbals. If they sound nothing like the real thing, well, you've got yourself a good chunk of playback distortion to work on! If your immediate reaction is, well, that's just a crappy recording, then Go Directly to Jail, Do Not Pass Go, ... :D - your first hurdle to overcome is the belief that crappy cymbals means a bad recording, rather than hearing distortion so obvious that it's difficult for the hearing system to work out what's going on - it sounds like a mess!

Having now got a good playback distortion sample, then focus on precisely how the quality of that distortion varies as changes are made to the system - you are now learning to how to distinguish changes subjectively ... :p, :D
 
Last edited:
You don't need them matched as you have a DC servo. I recon none will be too far away from you needs (build a jig) . The hiss I take very seriously . There can be no truth in FET square law curves being better here. Could my daft partition noise be a real thing? Could be one very big reason why JFET op amps do sound better? Also they are more RF immune. Doubtless others know this. I never looked it up so my best guess. If you think about it the current stream does part as in pentodes although not for the same reason. The JFET is just a flow with restriction. Bipolar is power steering and JFET a tap or faucet.

I still think you should have a current mirror . You will double your slew rate and have near symmetry. I presume to the VAS? If so doubly true as unlike me you don't use double LTP VAS? The latter forces balance through constant voltage due to two bases. Itself has a mirror to get the balance further up.

Are not the diodes to balance the VAS or force it to do something?

Surely we won't be arguing about matching being a good thing, are we, Nige?

As for the current mirror, I'm exploring my options on that front.
 
AX tech editor
Joined 2002
Paid Member
rare skills if true - how to establish their factual basis without testing with controls? - lots of anecdotes out there - to support just about any position

those who rankle at the idea of "self deception" being involved in expectation effects are denying basic human conscious perception - its not a choice, not about "honesty"

it is about how brains have evolved, how perceptual input is filtered/massaged/warped even suppressed way below the level of conscious awareness

Indeed. In the final analysis, 'you' (your conscious entity, your brain) is not at all interested in reality or truth or facts. 'You' are only interested in what advances you, is good for you, enhances your well being and survival.

Sometimes objective facts are useful, like knowing that jumping from the 6th floor will most probably kill you. Sometimes it is better for you to deny something that anybody else would call obvious.
It can be very subtle: if you root for team and they win, you'd say 'we won'. If they loose, you'd say 'they lost'.

As JCX states, it's all about being human. The unique property of the scientific method is that it tries to circumvent that human predisposition and thus advances the human condition. But the scientific method is pretty unnatural - there's books out there exploring what the survival value is of something that is so obviously unnatural and yet seemed to have been jumped up in existence.

Fascinating.

Jan
 
I think the FET thing is a false wall. There are sources for dual Jfets (Linear Systems for one) and while its not local to you its not impossibly far away or on the other side of a trade restriction. I'm sure there are people on this forum who could help. Otherwise selecting fets is not that hard. The Toshiba duals were selected singles clipped together.

Looking at an input differential in isolation is really difficult. The next stage will have a large influence on the details of the first stage. I would focus on CMRR first, otherwise the feedback stage will be dealing with it. Then on the common mode linearity. Anything you can do to reduce input modulations is really important since there is no mechanism to correct for those and they are more significant than you may think.

The version with the Zeners is bad news, Zener current noise will be the problem and why would you need them? Jfet/mosfet cascodes are great since they don't add noise (no base current getting added).


I think the JFET is a remarkable device. Please list some that are happy at 60 V if people know. It is a happy accident they will substitute a bipolar as the unsuitable MOSFET looks a better analogue in the generic LTP modern amp.

Dvv. I never used cascode LTP input. Will have to soon be it JFET or bipolar as my choices are running out.
 
Surely we won't be arguing about matching being a good thing, are we, Nige?

As for the current mirror, I'm exploring my options on that front.

In engineering the impossible must be made irrelevant. Matching can be best you can do in sensible time period. It isn't even difficult . BCV 61 and 62 are wonderful matches. Never do it yourself. Often their 20 V range is OK for driving a VAS (2 V max) . They even have a big pin to make it clear which is the diode side. Almost OK to hand solder even.
 
The problem with listening tests are we are not all to the same calibration in physical and nature/nurture thing. Thus we yield no useful data. The subtlety is like I give you AVO 8 volt meter readings and you a cheap Radio Shack meter. As we both use the same meter day in day out the experiments we do repeat. You with your Radio Shack might do better as you don't load down your circuits. We both need a Fluke meter and don't realize it. Problem is who is man enough to admit it? For all we know long lost friend Waveborn was the Fluke? He spoke as if he was. There was a guy with zero interest in transistors who knew more than us I suspect?
 
Status
Not open for further replies.