No. Noise is noise.
Unless you believe that everything is data. This would render the term useless, however.
As said before, it depends on the degree of a difference and a listeners ability to detect differences (under the specific conditions).
The way you want to collect the data you don´t know (in many cases like the one we are discussing in this thread) what differences were present in the samples, you don´t know enough about the test system and you don´t know next to nothing about the detection abilities of the listeners.
Didn´t that mean that you had to accept any test result as long as you didn´t get enough information to consider it as most likely "sloppy"?
For example, if i don´t tell you if a test was "blind" or "sighted" and furthermore as little as in this thread, wouldn´t you have simply to accept the results as data points?
Gathering of data is nothing to condemn, but that seems to be a strange approach where you get a lot of data points but don´t know what the independent and what the dependent variables were.
I´d interested to learn in which way any analysis of this data could be executed and what conclusions could be based on it.
DF96: In most of the threads where I have seen you post, you appear knowledgeable and to be making an effort to be helpful. However, I think you said one time that statistics was not your strongest subject. Perhaps that's why in you seems to be floundering in this thread. Instead of explaining the subject to everyone else, you seem be trying to find your own way. If I may, Jakob2 is doing what your usually do in most other threads, he is trying to educate those who lack proper understanding of the subject matter. In this particular instance it might be wisest to try to learn rather than teach. I mean that respectfully, as I have come to appreciate a lot of what you do, but I feel like some of what you have said recently is beneath your usual rigor and expertise. Also, probably best to avoid arguing with others who are in it to win more than to teach or learn, as it is unlikely to serve any productive purpose or reach any agreement congruent with proper science.
Last edited:
How come these "tests" rarely/never have staunch advocates from both sides fully participating? For instance I'm pretty sure Jan was outnumbered by valve folks at the Triode Fest with his 8-leg SS preamp and still came out on top.
What is the confidence level for your trials? You did not state that in your original post. I did not read over 100 pages of comments so maybe you stated it in there somewhere, but you need to make enough trials to establish that your hypothesis is correct, even if it's a negative result. ABX testing is only good if the confidence level is good, which requires many trials.
I'm perfectly in favor of ABX testing, and he did good to make sure the SPL was the same for each DAC. I'm well aware of psychoacoustics and the way that human hearing is easily fooled. An even tougher test, used in the food industry, is the triangle test. The only downside to ABX testing is the number of trials needed to achieve a 95% confidence level. An even better way to improve the test he did was to make every test subject wear a blindfold. Our eyes tell our ears what to hear, to a large extent at least. Any little visual sign of differences or colors or other visual things can have an effect on the outcome.
A better way to do this would have been with some good headphones, but he did good as far as I can tell.
A better way to do this would have been with some good headphones, but he did good as far as I can tell.
You might have added that somebody eventually comes along to insult as many people as possible in the fewest words.
Maybe a good idea to take a look at the forum rules from time to time, rule number 1 as I recall, is about disparaging other forum members. Violations can lead to consequences.
On the other hand, you would be welcome to participate if you have any thing on topic to say and you can do so without breaking rules.
Last edited:
It's not necessary to wear a blindfold if other means are available to keep participants from knowing which equipment is in use.
Also, don't know if the test was done blind or double blind. Double Blind would mean the experimenter, the OP I believe, also did not know which equipment was being listened to at any given time. In that case someone outside the room that the participants never saw would have to do the switching, or it would have to be done by automated randomizing switching equipment that conceals what is being listened to from everyone.
Last edited:
Same room same setup, alone in my case. Was 100% confident in my ability to discern both DAC's when doing sighted tests. I chose the tracks and duration. Can't find any pitfall in this test.
What differences did you perceive in sighted listening? Did you focus in on one particular sound element & isolate it in a portion of a track that you could easily identify between DACs when sighted or was it just a generalised difference?
Can you tell us about the general sound - was it expansive (soundstage width but particularly depth) & interesting or all the notes in the right place type of meh, uninteresting sound?
It might be interesting to do a blind preference A/B test - not trying to hear specific differences but listening as you would casually & choosing A or B as your preference. Enough trials to be statistically significant.
Mind you I would suggest not using optical cable as a first change to make in any new test
Last edited:
This is logically bankrupt.
There is no symmetry here - if the critics are biased it doesn't somehow counteract the obvious biases of the OP, and resuscitate the validity of the test. Any biases of the critics (or ardent supporters) are completely irrelevant as to whether or not the tests are useful or worthless. Neither the critics nor the ardent supporters have tests to defend.
It would seem that you agree with the "test" findings and will defend them, using whatever you can, including this sort non-sequitur.
Remember, the OP was confident that no one would be able to hear any differences in one of his "tests" BEFORE he even began it. His biases are clear.
that's gonna be so much fun to organize, this test...
audiocopy of a whole boutique system, which will include so much bulls*** mumbo jumbo AT THE SAME TIME
They'll be like:
''What do you think went wrong ?''
- ...ehm.. i think the Shunyata power cable was not quite broke-in yet
''oh no, you didnt!? I told you, Berthold! 1000 hours minimum!''
- I'm truly sorry, Oliver, *sobbing* i failed, it's all my fault!! *sobbing*
'' You're a shame for Audiophilia, Berthold, a bloody shame!!
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.