Resistor Sound Quality Shootout

[JMFahey]

what Daniel Kahneman knows: 'The human mind is a machine for jumping to conclusions.'

Of course, and that´s an important part of what makes us Humans.

 

[billshurv]

'The human mind is a machine for jumping to conclusions.'

And yet occam's razor never gets invoked here. Is it not exceedingly likely that the (now absent) OP wanted to believe a resistor in the least important place in the circuit would affect the sound and so it did?

 

[Markw4]

The OP said, 'Tests are still running, til now it's over 6 weeks now. 3 Persons are involved. Double blind tests and normal tests were used.'

Doesn't necessarily sound so simple as 'the OP wanted to believe...'

 

[Evenharmonics]

If I can't be there then I would at least like to pics of the experimental setup, a list of all the equipment used, and whatever other information is available.

Have you disclosed whatever other information of your experiments when asked?

Just don't like jumping to conclusions when information is sparse.

There are those who continue to do what they don't like. Must be some kind of business obligation...

There have been times in the past when I did jump to conclusions and later turned out to be wrong.

But you repeat it.

 

[john curl]

I want to expressly thank SIN GUN for posting his resistor listening results here on a public website. They essentially correlate with what we have found (subjectively) over the last 50 years, and early listening tests almost 40 years ago.

 

[john curl]

Here is an earlier cap and resistor (passive components) article:

 

[JMFahey]

I want to expressly thank SIN GUN for posting his resistor listening results here on a public website. They essentially correlate with what we have found (subjectively) over the last 50 years, and early listening tests almost 40 years ago.

Have you?
Wow!!!!!
Another idol falls down 🙁

 

[Jakob2]

Thanks for the PDF!

That was exactly Collom's article I've had in mind.

Wrt film caps used for coupling, we could find some evidence for audible differences in our very first single blind tests ~36 years ago. 🙂

 

[phase]

Maybe it’s more in the processing however, but somehow certain people fail to hear anything near the same things as others it would appear.

Cool Hearing Test: Are You a Superhuman? - YouTube

 

[billshurv]

best hearing test ever? Oh come on...

 

[billshurv]

The OP said, 'Tests are still running, til now it's over 6 weeks now. 3 Persons are involved. Double blind tests and normal tests were used.'

Doesn't necessarily sound so simple as 'the OP wanted to believe...'

Look at thread title 'resistor sound quality shootout'. I would say that pretty much predisposes the OP to thinking that different resistors have different 'sound quality' (whatever that actually is).



And note a double blind test required 3 people as minimum and only one of them is listening...

 

[kozard]

Yaego

Any idea how the following perform for distortion?

Stackpole RNF14
Vishay SFR25
KOA MF1/4

TyoHM
UNI-ROYAL


Any idea which is best and which is worst?


How does the Stackpole RNF14 perform compared to Dale RN55?

 

[mchambin]

And note a double blind test required 3 people as minimum and only one of them is listening...

I am not sure, the third one is listening, could very well be watching only.

 

[Zung]

... I referred to Ed Simon's measurements (see also his article, attached)...

Thank you for publishing the article.
A most interesting thread.

 

[johnmarkp]

Only two people are needed for this kind of test. One person installs the resistors and sets up the system. The second listens. This works as long as there is absolutely no interaction between the two. They could agree on a time schedule or some way of signally that the test is ready that does not involve direct communication.

 

[jan.didden]

If you look at the statistical concepts behind blind testing, you will find that you need a fair number of participants to get the confidence numbers up into 80% or 90%.
A controlled test with a single participant basically doesn't give you anything in terms of confidence that the results are indicative for the actual differences or not.
That's one of the reasons why well designed controlled tests are so rare. Apart from the mechanics, it's not easy to get, say, 10 experienced listeners to put in the time to participate.

Jan

 

[billshurv]

Only two people are needed for this kind of test. One person installs the resistors and sets up the system. The second listens.

That's not double blind, which the OP said he used.

 

[Jakob2]

If you look at the statistical concepts behind blind testing, you will find that you need a fair number of participants to get the confidence numbers up into 80% or 90%.
A controlled test with a single participant basically doesn't give you anything in terms of confidence that the results are indicative for the actual differences or not.

Simply "no", the statistical analysis does not care about the number of participants it depends only on number of answers overall and the number of correct answers.
If for example 36 correct answers are provided by one participant or by ten doesn't matter, if the hypothesis to be examined is whether an audible difference for a certain effect exists.

Wrt conclusions from test results, it might be different.

 

[Jakob2]

That's not double blind, which the OP said he used.

Information about what was done is too sparse for any judgement, but basically two people could do a "double blind" test.

 

[mchambin]

Valid results from a blind test need more than one listener because of statistics. This is well explained in post #256.
Furthermore, in "double blind", the "double" means the guy who switches the stuff doesn't know the nature of what he is swapping.