Thread split from - Blowtorch part II______________________________________________________
Is this not a contradiction in terms - emphasis is mine?
Not at all.
Most things have some flaws. If these are understood they can be accounted and allowed for in drawing conclusions.
If we waited for infalllable tests for everything, we would still be waiting.
Science, by definition is not infallible, so that argument is a straw man one. What's your definition of reliability then?
Soundminded's paragraph before my quoted sentence
Tell me where you draw the line to decide when a test is "good enough"?
but when related evidence suggests the test flaws have more effect on the test outcome than the claimed hypothesis we do need to address those flaws
in subjective testing of human perceptions it's pretty clear that sighted tests have flaws too big to be relied on in many situations
on the other subject engineering principles can be used to translate measurements of an effect in one system to predict its impact on the perfromance in another circuit topology
tape heads have air gaps, different windings than purpose designed MC step up xfmr - so the papers/discussions Mr Curl mentions should be presented in/translated to "transferable" engineering models - noise source levels and impedances, mutual inductance factor estimates may be needed to adjust for winding differences
as mentioned in the Stereophile phono EQ article 2nd order roll off (like xfmr hf response) can have much less amplitude error than 1st order filters over some frequency range
curious minds do wonder why we're seeing seemingly randomly chosen, poorly contextualized objections rather than comments on measurements, datasheet numbers of existing MC xfmr for the bandwidth, flatness issues, or valid information from the cited sources translated to the conditions of the MC xfmr circuit model
an engineering approach to me means use good models, measurements and engineering knowledge, models to calculate the magnitude of effects like lamination thickness - in the intended application: MC step up xmfer
if it is a "problem" by whatever criteria we choose perfromance "requirements" then you can see what solutions are technically available - "tape wound" "bobbin" cores can be had with 1/8 mil Supermalloy
in subjective testing of human perceptions it's pretty clear that sighted tests have flaws too big to be relied on in many situations
on the other subject engineering principles can be used to translate measurements of an effect in one system to predict its impact on the perfromance in another circuit topology
tape heads have air gaps, different windings than purpose designed MC step up xfmr - so the papers/discussions Mr Curl mentions should be presented in/translated to "transferable" engineering models - noise source levels and impedances, mutual inductance factor estimates may be needed to adjust for winding differences
as mentioned in the Stereophile phono EQ article 2nd order roll off (like xfmr hf response) can have much less amplitude error than 1st order filters over some frequency range
curious minds do wonder why we're seeing seemingly randomly chosen, poorly contextualized objections rather than comments on measurements, datasheet numbers of existing MC xfmr for the bandwidth, flatness issues, or valid information from the cited sources translated to the conditions of the MC xfmr circuit model
an engineering approach to me means use good models, measurements and engineering knowledge, models to calculate the magnitude of effects like lamination thickness - in the intended application: MC step up xmfer
if it is a "problem" by whatever criteria we choose perfromance "requirements" then you can see what solutions are technically available - "tape wound" "bobbin" cores can be had with 1/8 mil Supermalloy
You have misread what I said or at least meant to say. The fact that a particular test or set of tests are flawed does not change the fact that testing is the only path to knowledge. That is an inherent part of the scientific method that is distinguished from natural philosophy where people just contemplated their navels to learn the truth. That's what led them to believe that everything was made out of earth, water, fire, and air. If the test is flawed, devise a better test, don't give up on the notion that testing is crucial. The scientific method consists of hypothesis, testing, results, and conclusion. If the test is inadequate then the conclusion will be invalid. But it is only proven methodology for getting at the truth.
If the market is ignorant and unregulated, sell them snake oil, they won't know the difference. No need to bother with engineering in that case, just as long as it tastes and smells foul, they'll think they're getting better from it.
Studying a bit of philosophy might be a good way of avoiding the logic flaws in your argument, perhaps?
Oh boy!
There are things in life that can be known by experience, rather than by testing.
Love is one example. Elated spirit that stems from viewing some nature views is another.
When it comes to the joy of listening to music, I choose experience and skip testings entirely.
Audio replay is an illusion - you are not approximating the real thing - you're creating a more pleasing illusion - nobody can measure that scientifically.
A sound system's proximity to the source is only partly measurable. Especially data published by manufacturers have very little meaning as for the sound quality.
I'm looking to meet even single serious music listener who choose loudspeakers by price and measurements only, without listening to them.
I have a feeling that all those who applaud measurements only don't act accordingly when it comes to choosing loudspeakers for themselves.
You might want to look at that paper mentioned in post #11062. I think we agree too. A coil over a metal object exchange energy at equilibrium, therefore an eddy current loss genenerates heat and that is the only mechanism for the noise. Excess noise is a particular non-ideal property of some matierials and I am back to disagreeing with jneutron (though I understand it is at my peril).
I almost never know the original, but I have some good experience of the sound of real instruments and the human voice. I acknowledge it is unlikely that we will ever approach perfect reproduction...that does not negate the huge progress we have made on the back of solid technical principles and measurements in more closely approximating that goal. You surely can't believe we would have reached the current state without the benefit of this technical, measurement based approach?
All sensory perception is an illusion, even live music. It's not reality-in-itself that enters our minds, it's the sensory-mental reflection of reality which is perceived by our minds.
However, when it comes to reproduced music, some systems create in the listener a perception which is closer to the perception of live music, closer than other systems. AFAIK, there is no music reproduction system that creates identical perception to live music. The question is how far, or how close, a system is, compared both to live music and to other systems.
No, one could not.
You are taking my saying out of context.
An audio designer should be well versed in known knowledge and should also recognize what can be measured and what should be experienced (by listening tests).
Actually an illusion is a distortion of sensory perception that reveals the underlying mechanisms of perception. As such normal sensory perceptions are not illusions, but constructs the mind uses to integrate the senses into an inner reality that mirrors the outer reality. If this were not produced with very high fidelity, we would not have made it here to have this exchange.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.