Blind Listening Comparison

[rDin]

Go for it. Sounds like fun. Love doing these sorts of tests. You may learn something. You may not.

The only comment I'd add, based on my experience, is that while it may be easier to discern differences, it's often harder to determine a preference. If differences are noted, you may then want to spend more time with each to determine if there is a preference.

 

[Gregadd]

Just remember this anthem of the repair industry- " You buy cheap. You by twice."

 

[Gregadd]

Go for it. Sounds like fun. Love doing these sorts of tests. You may learn something. You may not.

The only comment I'd add, based on my experience, is that while it may be easier to discern differences, it's often harder to determine a preference. If differences are noted, you may then want to spend more time with each to determine if there is a preference.

That is why I suggest developing your preference before you do the test. If there is no preference then you are done. Then when you do your blind test you will know what to listen for.

 

[microstrip]

Go for it. Sounds like fun. Love doing these sorts of tests. You may learn something. You may not.

The only comment I'd add, based on my experience, is that while it may be easier to discern differences, it's often harder to determine a preference. If differences are noted, you may then want to spend more time with each to determine if there is a preference.

Can you explain with execution details and analysis the listening blind tests you carried?

Although people can have a good and fun time carrying them, they also risk getting the wrong results from them and take the wrong "unquestionable" blind decision.

 

[Gregadd]

I would be willing to wager that your results will be Inconclusive (consistent with chance). That is what blind test advocates hope for.

 

[wil]

Can you explain with execution details and analysis the listening blind tests you carried?

Although people can have a good and fun time carrying them, they also risk getting the wrong results from them and take the wrong "unquestionable" blind decision.

Are you saying blind tests are more likely to result in perceptual errors than sighted listening comparisons?

I think both methods are equally fraught if not done over a unhurried period of time and in a relaxed frame of mind.

Given the choice, I’ll choose blind. Expectation bias can have a powerful influence — and all the more so because we don’t know it’s happening.

In my jaded opinion, a significant portion of the audiophile biz exploits and depends on this vulnerability to bias.

 

[Gregadd]

I find there to be significant exploitation on both sides. On the blind test side is the assumption that the null hypothesis has been proven because of statistically "insignificant" results. On the other side there is to much hyperbole related to the proposed improvements. Moreover the discover of the problem is also the seller of the cure.
Uh Oh, There i go down the rabbit hole.

 

[NC Lee]

I started typing answers and got overwhelmed, so, a bit more info for now:
I A/B'd quickly and I heard clear differences; Did again the next day with same result. Wife heard same differences both times though she didn't know which cable was which.
The differences seemed so clear, the double blind test will be for fun as another check on just how clear the difference is, or to validate such.
I can quickly and always tell a trumpet from a piano. Can I always hear the fullness in one cable v other?
I agree with Elliot. Then I'll listen to the new cable for a week and change back to the prior cord. Past equipment comparisons, I'd hear a difference immediately and want its strengths/preferences, only to change my mind on longer listening, not willing to accept the lessened aspects.
Here, I'm not seeking statistical significance just another data point. And learning more about myself and my system.

 

[Amir]

Our hearing mechanism is so complex and our ears can train and receive more information.
I think for Test and comparing A vs B the time is important. The Brain can learn more across the time and I prefer to start listening to A for hours and hours to be sure my brain received enough information.
After listening to A I switch to B and start listen carefully for hours or days.
Then I start the blind test and ask someone to help.
I ask him to repeat the test for more than 6 times and If I detect correctly all times then I am sure the test result is Valid.

Be sure With Simple measurements and without modeling hearing system no objectivist could attack our hearing result.
Both Audio systems and Hearing/brain Systems are non-linear complex systems and objectivist modeling/measurements are all on simplified models.

 

[Amir]

Detecting A vs B is not hard task for a listener but the Important question is "which is better?"

Right answering to this question needs:
- high level of experience (expert audiophiles)
- reference/match system
- properly setup system

if you test a dac (for example) in a system what you hear is not only dac , you hear also all problems of system and you may prefer a low performance dac to high performance dac in a low performance system.

I know audiophiles with more than 30 years buying selling upgrading attending shows and ... but they never had right judgments and never get a good sound in their home.

Here the experience of expert audiophiles like Jim Smith will help us.
Audio is not a simple game

 

[Gregadd]

[1]Detecting A vs B is not hard task for a listener but the Important question is "which is better?"

[text omitted]

[2]I know audiophiles with more than 30 years buying selling upgrading attending shows and ... but they never had right judgments and never get a good sound in their home.

[1]Here the experience of expert audiophiles like Jim Smith will help us.
Audio is not a simple game

1.IME it has been a pretty demanding task A precious few have been successful.
. It is difficult to be wrong in a preference test.
2. It must have been difficult for you to not point out the emperor had no clothes.
3. Expert audiophile. I suppose one can be an expert and an audiophile. I am not sure what an expert audiophile is.
Being am audiophile is easy to do. difficult to master.

 

[microstrip]

Are you saying blind tests are more likely to result in perceptual errors than sighted listening comparisons?

Still waiting to read a report of a proper blind listening test carried by audiophiles with high-end equipment.

I think both methods are equally fraught if not done over a unhurried period of time and in a relaxed frame of mind.

Yes. But most people manage to carry sighted tests for long time, never heard about a long term blind test.

Given the choice, I’ll choose blind. Expectation bias can have a powerful influence — and all the more so because we don’t know it’s happening.

Do you want to report us of such tests and their outcome?

In my jaded opinion, a significant portion of the audiophile biz exploits and depends on this vulnerability to bias.

We fully agree on this. But IMO even with bias we educate our preferences and become happy and knowledgeable listeners!

 

[rDin]

Can you explain with execution details and analysis the listening blind tests you carried?

This question feels loaded? I'm not a scientist, so any method I use is unlikely to satisfy scientific rigour, but it should be recognised that the validity of the results are not solely dependent on the method used. if the results are testable and repeatable, then to dismiss them due to lack of scientific rigour would be rather short-sighted?

 

[Gregadd]

"I decided to not post some snarky comment.
"Those who seek the chaperone of science need some science."

 

[Tim Link]

I would be willing to wager that your results will be Inconclusive (consistent with chance). That is what blind test advocates hope for.

Not this blind test advocate! It's only when differences can be detected blind that things start to get interesting. At that point we can try making some measurements. If the blind test successes remain but we can find nothing in measurements, then that would be super exciting!

 

[Gregadd]

Not this blind test advocate! It's only when differences can be detected blind that things start to get interesting. At that point we can try making some measurements. If the blind test successes remain but we can find nothing in measurements, then that would be super exciting!

Can you cite an example when that happened. I will accept anecdotal evidence for purposes of discussion.

 

[Tim Link]

Can you cite an example when that happened. I will accept anecdotal evidence for purposes of discussion.

An example I read about recently was from Benchmark Audio. I'm not sure they double blinded it but they were hearing distortion that wasn't showing up on their measurement tests, at least not until they heard it and started looking for an explanation. Their explanation was inter-sample overs - a term I had never heard before. These occurred during some music playback but apparently not from their standard test signals. That's my understanding. They talk about the topic here: https://benchmarkmedia.com/blogs/application_notes/intersample-overs-in-cd-recordings
It seems to be a potential problem with digital recordings that may explain at least part of what people are claiming to hear that they don't like about digital. On my own system I've had plainly audible distortion crop up, making me think I had a rubbing coil or something. It was plainly obvious so I didn't test it. My friend heard it start up too so it wasn't just me. Strangely I wasn't getting anything on a sweep measurement. It either played the sweep fine, or the analysis software filtered out the problem as background noise. I couldn't hear it on the sweep. With music it almost sounded like someone playing along faintly with a kazoo until I restarted everything.
More recently I've been hearing what seem like occasional skips in streaming music, like a record skipping. So I restart the stream. Another weird thing I've heard is wow and flutter - really bad wow and flutter. On a pure digital signal path! Something in the signal chain had changed to a different sample rate when I used Garage Band and for some reason the playback speed was varying to keep two sample rates in sync. instead of just doing a sample rate conversion along the way. I didn't try running a sweep with that one. Might have been interesting! If these really obvious problems can occur with digital I don't doubt some more subtle ones may occur too, and so there could be some differences in digital equipment that seems to all measure "perfectly" that would show up in a blind test.
Another thing I heard and had verified was with a Moodlab dac I bought years ago. I could hear distortion. After doing some research I learned that the power supply they had provided didn't have enough current to properly drive the dac. I found a higher current power supply and the problem went away. I didn't test that but I wouldn't be surprised if it didn't show up on a sweep every time. It would distort several times a minute.

 

[microstrip]

This question feels loaded? I'm not a scientist, so any method I use is unlikely to satisfy scientific rigour, but it should be recognised that the validity of the results are not solely dependent on the method used. if the results are testable and repeatable, then to dismiss them due to lack of scientific rigour would be rather short-sighted?

The question is that you are playing with words and concepts. You want to use blind tests because they are the scientific way to carry listening tests without bias and then use unknown methods that hide behind secrecy .
Science also uses common science, but is always public and scrutable.

This is an hobby, anyone can live it its own way. But IMO someone wanting to talk about listening bias and blind tests must be able to understand them.

My main point - unless they are well understood and properly carried blind tests are even worse than listening bias.

 

[steve59]

I read the op then skipped to the end. slippery slope, a local dealer told me a story about power cables involving a customer calling and telling him his new cord transformed the soundstage from his stereo. Mike, the shop owner wen on to say he now had to wait 24 hours for the amp to return to running temp b4 he could do a fair comparison also telling me he was as familiar with this mans system as he was his own. To this I got nothing.

You sir do have a keeper in that wife of yours tho’

 

[rDin]

The question is that you are playing with words and concepts. You want to use blind tests because they are the scientific way to carry listening tests without bias and then use unknown methods that hide behind secrecy .
Science also uses common science, but is always public and scrutable.

This is an hobby, anyone can live it its own way. But IMO someone wanting to talk about listening bias and blind tests must be able to understand them.

My main point - unless they are well understood and properly carried blind tests are even worse than listening bias.

You're setting up a false dilemma - either it's the "properly carried blind test" or it's failure.