John Curl's Blowtorch preamplifier

Status
Not open for further replies.
Jump to Latest
#12,541
I believe we lose low level detail in all equipment, some only more than other. What measurement will show this?
Click to expand...

Distortion and noise. Either the DUT is faithfully replicating the input signal or it isn't. This is a rather simple electrical measurement and one done routinely. For the example you gave earlier (the presence of bass causes the treble to blur), one would look for intermod and PS modulation products in the output spectrum. This isn't rocket science.
 
#12,543
Jakob2 said:



I´m terrible sorry janneman, but your´re not right in this point.

This sort of test is called ´forced choice test´ and is well established in sensory tests.

A test with two possible answers `a better than b´and ´b better than a´ is called a 2-AFC-test (two alternate forced choice), if you do such a test with a group of listeners listening at the same time, it can be better to include a third choice `no difference´to avoid rumors in the panel from listeners hearing no differences. That would be a 3-AFC-test.

Wishes
Click to expand...

Hi Jakob,

As far as I know, a force test is done to establish preferences between two clearly different products. For example, whether people would prefer an Arctic Gray car color over a Solar Gold car color. I don't think that is appropriate when yoiu want to find out if there are (audible) differences in the first place. But I am not very familiar with that type of test. Do you have a link to more info on this?

Jan Didden
 
#12,546
janneman said:


Hi Jakob,

As far as I know, a force test is done to establish preferences between two clearly different products. For example, whether people would prefer an Arctic Gray car color over a Solar Gold car color. I don't think that is appropriate when yoiu want to find out if there are (audible) differences in the first place. But I am not very familiar with that type of test. Do you have a link to more info on this?

Jan Didden
Click to expand...

For a short explanation, just take a look here:

http://www.tonmeister.ca/main/textbook/node361.html

A more detailed example for the usage of forced choice tests is this:

http://www.ak.tu-berlin.de/fileadmin/a0135/Magisterarbeiten/Karl_Roman_Magisterarbeit.pdf

(Audibility of signal reflections, Magisterarbeit), unfortunately written in german.

A lot of books for sensory evalutation techniques cover this topic.

It sometimes isn´t so obvious, but double blind testing removes only one certain type of bias, all other possible forms of biases remain unadressed and the experimentator has to take them into account.

Wishes
 
#12,547
Andre Visser said:


Stinius, I believe the things we discuss are a bit more complicated than 2+2=4.

Perhaps you are right about the 'too educated' part, luckily not all engineers think they know everything and are still willing to explore the unknown.
Click to expand...
exploring the unknown... that's a scientist.
engineers stick to applying what is already known.
 
#12,548
Andre Visser said:


Stinius, I believe the things we discuss are a bit more complicated than 2+2=4.

Perhaps you are right about the 'too educated' part, luckily not all engineers think they know everything and are still willing to explore the unknown.
Click to expand...


janneman said:



Edit: What about the ABX test? You, as the listener, have a switch and you can listen to amp A, amp B, or amp X, where X is either A or B, but you don't know (nobody knows, even not the test director, and this is critical).
Your task is to identify whether X is A or B. Of course X is changed random everytime you select it.
I don't think it can get simpler than that. You can switch whenever you want, to A, B or X whatever you want. You can play the music you want, for as long or as short as you want.
If you cannot correctly identify whether X is A or B, in a statistically significant number of trials, I'm willing to accept that there is no audible difference between A and B.

(Let's for the sake of discussion assume that the test box itself doesnot mask the difference. It is a valid issue, but let's just concentrate on the concept here. One step at the time).

Jan Didden
Click to expand...


I would be willing to acknowledge that *I* do not hear a difference (under the specific test parameters) but I would not go so far as to say no one, ever can hear a difference.
The test is not valid to say no one ever.
 
#12,551
myhrrhleine said:






I would be willing to acknowledge that *I* do not hear a difference (under the specific test parameters) but I would not go so far as to say no one, ever can hear a difference.
The test is not valid to say no one ever.
Click to expand...


Well, if you replace the "I" by a large number of listeners doing a large number of trials, you can say that you are, for instance, 99,99% sure that there is no audible difference. Of course, at one day somebody may turn up who scores 95% right in the test, and that then proves that there is indeed a difference. But that difference is then so subtle that only 1 listener in 1000 can hear it.

For a manufacturer it will probably not enough to put money into that difference. Much better to put money in some marketing to convince the other 999 that there is a large difference that you can't miss, unless you are deaf! 😉

Jan Didden
 
#12,552
Joshua_G said:



Is bias being distinguished from personal preferences?
Click to expand...


I don´t know if i got your question right, but my answer would be yes.

But personal preference can cause a bias.

Just for example, if you prefer in general tube amplifiers, you might rate the sound quality higher if you see that a tube amplifier is used.

Wishes
 
#12,553
Jakob2 said:


For a short explanation, just take a look here:

http://www.tonmeister.ca/main/textbook/node361.html

A more detailed example for the usage of forced choice tests is this:

http://www.ak.tu-berlin.de/fileadmin/a0135/Magisterarbeiten/Karl_Roman_Magisterarbeit.pdf

(Audibility of signal reflections, Magisterarbeit), unfortunately written in german.

A lot of books for sensory evalutation techniques cover this topic.

It sometimes isn´t so obvious, but double blind testing removes only one certain type of bias, all other possible forms of biases remain unadressed and the experimentator has to take them into account.

Wishes
Click to expand...


Danke Jakob, ich habe keineswegs Probleme mit Deutsch lesen und verstehen!

Jan Didden
 
#12,556
Jakob2 said:



I don´t know if i got your question right, but my answer would be yes.

But personal preference can cause a bias.

Just for example, if you prefer in general tube amplifiers, you might rate the sound quality higher if you see that a tube amplifier is used.

Wishes
Click to expand...


My personal preferences are transparent, detailed, accurate, neutral and natural sounding audio system, including the soundstage.
 
#12,559
janneman said:



Well, if you replace the "I" by a large number of listeners doing a large number of trials, you can say that you are, for instance, 99,99% sure that there is no audible difference. Of course, at one day somebody may turn up who scores 95% right in the test, and that then proves that there is indeed a difference. But that difference is then so subtle that only 1 listener in 1000 can hear it.

For a manufacturer it will probably not enough to put money into that difference. Much better to put money in some marketing to convince the other 999 that there is a large difference that you can't miss, unless you are deaf! 😉

Jan Didden
Click to expand...

Supposedly, this thread is about the blowtorch.
It isn't for the 99.99% that can't hear the difference, but rater that small group that can.
 
Status
Not open for further replies.
Top Bottom