That's not strictly true. You measure it with properly conducted DBLTs.
Then you can say stuff like, "99% of listeners can't hear this reliably". From nearly 2 decades of DBLTs, it's fairly easy to spot da Golden Pinnae listeners who are deaf. eg those who hear differences between mains cables. So it's true we don't all hear the same.
But I was shocked that some of my DBLT panel could, under certain circumstances & on certain test signals, reliably distinguish frequency response variations of 0.1dB 😲 At that time, I couldn't conveniently measure response to better than 0.2dB
Is that important? Only when you are conducting DBLTs on stuff like phase distortion, dither bla bla. For speech & music, even my panel, which included some of the best ears in the business, was happy with 0.2dB flatness and that is still my target for most stuff ... though Guru Scott Wurcer took me to task for having such low standards 😊
Are you feeding a 1Vpp squarewave into the RIAA preamps ? Or is it 2Vpp (+-1Vp) ?
What are the 'Transition' waveforms?
hbtaudio, I don't object to your manner which I rather like. But you haven't answered my questions.
Hans,
I'm watching this thread with interest. Repeatedly one article by late Graham Maynard popped up in my mind (no pun intended). If this was posted before, I'm very sorry, couldn't find. To my liking he went very practical and "anti audiophool" with this one, and accompanied it with -kinda- logic commentary, but without ANY measurements, nor simulations. So I will not defend or explain his ideas. Just for fun (which sadly periodically disappear from the discussion). So here you are:
https://bovan.net/gmweb2/the_ace.htm
Btw I didn't build it yet. But bought components already.
https://bovan.net/gmweb2/the_ace.htm
Btw I didn't build it yet. But bought components already.
Attachments
@kgrlee
I assumed that basically everything is self-explanatory, but in a nutshell the answer is:
Amplitude 1V (no DC component) = 2 volts peak to peak
For the periodizing (square, not sinc or dirac ...) pulse 0V (DC reference potential) jumps to n volts ---> +2V.
In both examples (the pure cascaded filter function and a single-stage dc-coupled allInOne RIAA-EQ with G1k = +40dB) there is no high-pass or DC blocker. So no differentiating effect is possible.
The (red) Y(t) labeled “Transition” start at zero, so these are the transient analyses.
If there is a need for further clarification, please send me a PM.
All the best,
HBt.
Final addendum
The purpose was simply to illustrate that there can be no problems with clicks and pops for the electronics and the underlying filter function.
HBt.
The purpose was simply to illustrate that there can be no problems with clicks and pops for the electronics and the underlying filter function.
HBt.
There is absolutely no reason to get upset. Assessments of a situation always differ. Weightings are not automatically set equal.
As long as the black box does its job, which is mainly to amplify and equalize the tiny signal, everything is fine.
Whether I'm driving a TESLA or a CHRYSLER, a VW Golf or an Audi Q7 doesn't change anything:
the vehicle gets me from A to B.
#
Emotions are out of place in any relevant, technical discussion.
It is pure deduction based on logic - and knowledge that has already been acquired.
I wish us all a peaceful Sunday,
HBt.
As long as the black box does its job, which is mainly to amplify and equalize the tiny signal, everything is fine.
Whether I'm driving a TESLA or a CHRYSLER, a VW Golf or an Audi Q7 doesn't change anything:
the vehicle gets me from A to B.
#
Emotions are out of place in any relevant, technical discussion.
It is pure deduction based on logic - and knowledge that has already been acquired.
I wish us all a peaceful Sunday,
HBt.
Any form of testing requires a proposition as to the cause of a perception, hence only the proposition is tested, being refuted or otherwise. For someone to declare "I heard a difference" cannot be refuted as real by a faulty proposition of the cause being tested.
Value attributed to "I heard a difference" has variant merit dependent upon the object, one that can be personal. This as oftentimes to take some action as a result. DBLTs have value that isn't solely or globally relevent to all things imagineable.
I generated some clicks and pops so you can hear for yourselves. https://www.diyaudio.com/community/threads/take-the-click-and-pop-challenge.424691
Ed
Ed
reference #212
Hello Hierfi,
for exactly this reason, a huge number of additional questions are always asked - to enable an evaluation in the first place - and must be answered.
If we are then able to identify two or more typical representatives (electronics) with statistical certainty, only then do we proceed to the next round.
A further list of questions is followed' much later by an initial preliminary measurement phase.
At some point, we finally come to a green branch.
kindly,
HBt.
Psst
The first thing we need for examinations is two or more pobands. Without concrete samples we are powerless.
Ethan and Paul
I'm 100% on Ethan's side, but I can also categorize Paul's statements - so from that point of view there's no reason for a petty war.
Explanation:
The video link should be understood as illustrating what we are discussing here in this and countless other threads and often arguing about with an emotional attitude.
Hello Hierfi,
for exactly this reason, a huge number of additional questions are always asked - to enable an evaluation in the first place - and must be answered.
If we are then able to identify two or more typical representatives (electronics) with statistical certainty, only then do we proceed to the next round.
A further list of questions is followed' much later by an initial preliminary measurement phase.
At some point, we finally come to a green branch.
kindly,
HBt.
Psst
The first thing we need for examinations is two or more pobands. Without concrete samples we are powerless.
Ethan and Paul
I'm 100% on Ethan's side, but I can also categorize Paul's statements - so from that point of view there's no reason for a petty war.
Explanation:
The video link should be understood as illustrating what we are discussing here in this and countless other threads and often arguing about with an emotional attitude.
"for exactly this reason, a huge number of additional questions are always asked - to enable an evaluation in the first place - and must be answered."
This presupposes that you get to ask questions of those stating "I heard differences". A failure to respond does not refute the reality.
Further, what guarantee is there that questions and answers are relevant to something unknown in the creation of a legitimate proposition in the testing of something unknown. This requires an assertion of a complete body of knowledge.
This presupposes that you get to ask questions of those stating "I heard differences". A failure to respond does not refute the reality.
Further, what guarantee is there that questions and answers are relevant to something unknown in the creation of a legitimate proposition in the testing of something unknown. This requires an assertion of a complete body of knowledge.
Last edited:
A properly conducted DBLT is designed to address exactly this. There are 2 propositions tested. The first is "I heard a difference".
You do this in an ABC test by having 2 of the presentations the same. Do the test 3x. If the victim picks the same DUT each time and says 'similar' things about them, there is a very high probability he can hear a difference.
BTW, you NEVER tell the victim what he is listening to ... or that 2 are the same. He might be listening to 3 different mains cables as far as he is concerned.
If and only if this test is positive, do you ask other stuff like, "Which do you prefer?" .. which is the second 'proposition'.
You discount the opinions of those who don't get the 3 tests 'correct'. For these deaf Golden Pinnae, just tell them very loudly & clearly, "My stuff is hand carved from Unobtainium & solid BS by Virgins".
Apologies to da statisticians among you for my crude pontificating on 'sparse data'.
The proposition "I heard", in the declaration "I heard a difference", is a past event that cannot be tested without the creation of a test mechanism that addresses the nature of the "hearing" of a past reality. Hence for "properly conducted DBLT" testing to occur in addressing a claim "I heard a difference" is not possible without a prerequisite assertion that all is known, whereupon the "cause" for the hearing can be tested, as going back in time isn't possible.
My argument isn't that DBLT has little or no merit, rather to challenge that DBLT is faultless. Untested or untestable claims of "I heard a difference" cannot be dismissed being false or lacking value in absence of DBLT testing. This does not prevent the weighting of untested or untestable claims to the extent they can be regarded valueless to variant objects, as also leading to questioning the objectives in determining value toward those objectives.
My argument isn't that DBLT has little or no merit, rather to challenge that DBLT is faultless. Untested or untestable claims of "I heard a difference" cannot be dismissed being false or lacking value in absence of DBLT testing. This does not prevent the weighting of untested or untestable claims to the extent they can be regarded valueless to variant objects, as also leading to questioning the objectives in determining value toward those objectives.
What you are saying is that the event cannot be replicated. If so, it can't be tested per scientific method bla bla.
But also it means the proposition is of no consequence to a maker or DIYer as it cannot be replicated. For this, the correct course of action is to say, "My stuff is hand carved from Unobtainium & solid BS by Virgins" 😊
If it can be replicated, it can be tested though this might be difficult and expensive.
BTW, besides finding out who da deaf Golden Pinnae are in DBLTs, one thing you WILL find is that the Woman in the Street is usually more perceptive (gives more reliable and repeatable results) in DBLTs than da Man in the Street who is usually more perceptive than HiFi Reviewers.
So pay attention when your wife, girlfriend, mistress or mother says, "I don't like this new one as much as your old one". 😲
"What you are saying is that the event cannot be replicated"
No. You cannot replicate what is not known of a past event as the potential cause of a declaration "I heard a difference" to conduct a test of that unknown. What are you testing for?
No. You cannot replicate what is not known of a past event as the potential cause of a declaration "I heard a difference" to conduct a test of that unknown. What are you testing for?
All is now clear. The 2 Vpp single sided Square wave will certainly severely overload the 40dB@1kHz RIAA stage with 5532/4 at 1kHz, 'slightly' overload at 8kHz and probably not overload at 16kHz which just shows the HF filtering of RIAA.
Da 'transition' or transient waveforms show the start of these with the output moving up cos the DC in the single sided input.
hbtaudio, I'm replying here instead of PM cos I think others with a small brain like me would like an answer too
I was puzzled by the '+1' references but now think it is the Supersonic pole required for absolute accuracy in a non-inverting feedback RIAA as per Lipshitz bla bla. Could someone enlighten me?
If so, I have conducted BLTs (not Double Blind) on my last commercial RIAA preamp in Jurassic times. It originally had the supersonic pole in a 40dB@1kHz stage like Bonsai. I got rid of it as a result of this BLT and also cos the response was still 0.2dB accurate @ 20kHz
Don't hit me Bonsai for not designing stuff properly 😲
BTW, BLTs are sufficient if the result is always 'NULL' ie no one can tell any difference. But if one of the true golden pinnae in my DBLT panel says they can hear a difference, we have to go on to the HUGE & $$$ trouble of DBLTs.
Last edited:
"All is now clear. The 2 Vpp single sided Square wave will certainly severely overload the 40dB@1kHz RIAA stage with 5532/4 at 1kHz, 'slightly' overload at 8kHz and probably not overload at 16kHz which just shows the HF filtering of RIAA."
Likely 2V before the inverse RIAA network.
Likely 2V before the inverse RIAA network.