[DMF]

Craigsub will be auditioning two speakers he's never heard before - one NIB (new in box), the other broken in for 200 hours. Speakers are two-ways, driven by his extremely competent gear.


Predict whether he'll hear any difference that can be ascribed to a "break-in period".


Manic discussion thread is here .

 

[DMF]

That's a "fatal flaw".


I did forget one, though: The "what the heck are you talking about / CowboyNeal" option.

 

[Rijax]

I think you should've done it this way:


I believe speakers break in.


I believe speakers AND ears break in.


I don't believe speakers break in.


I don't believe speakers break in, but ears do.


I don't believe speakers OR ears break in.


I haven't got the foggiest!


After all. Who can predict what a looney like Craig is gonna hear?

 

[danbry39]

As I said in the other thread, there are other gradients between "big" and "no" difference. The results of the poll ought to be interesting though...as will Craig Subs findings of his little project.

 

[DMF]

There is an option between the "Big" and "No" options. See it now? Even if you don't read any French, the English phrase "Only a certain" should be a clue.

 

[DMF]

Rixax, that's a fair poll. Wanna start it?

 

[ptaaty]

I think Craig will hear a difference with the Onix Ref speakers.


Previously I had not put much stock in break in (other than for subs, which have large mechanical movement, they will play louder and sound different usually after break in).


Then I head the Onix Ref 2s (at Kris Deering's, who was reviewing them), the Vifa Ring tweeter was harsh and grainy, and at first thought who would ever buy these speakers.


I went home to my Axiom system and was pleased with the highs, not nearly as grating. (I know what you are thinking!)


Then two weeks later after Kris had been using the Ref 2s for awhile I repeated the same listening test...this time the high end sounded much better. It still didn't have the very top in (+15kHz?) that I think adds some "air" or presence, but the sound quality was much better. In fact they now sounded better than the Axioms (which Kris had also reviewed and I heard on the same system).

 

[tonygeno]

The test is flawed. If he hears a difference, there's no way of knowing if the difference existed even before the break-in of one of the speakers as he never compared the two brand new speakers.


If break-in takes 100's of hours then a 30 minute initial test would not have affected the results. If break-in occurs in 30 minutes, then who cares?

 

[tonygeno]

Quote:

Originally posted by craigsub So it is your position that the Ref 2's are not a well matched speaker, got it.

Craig:


That is absolutely NOT my position. Let me explain it in more detail:


Speakers that are matched within +/- 1/2 db would be considered very well matched. About as state of the art as you get today (although I know there are some esoteric brands that claim even closer pair matching). Hopefully, the Refs are matched to these tolerances, which would be fabulous. But here's the rub: if in the midrange one speaker is going up that 1/2 db and the other is going down, the difference will be 1 db at those critical frequencies and most likely audible, (particularly if the deviation is broadband). Now, in stereo listening with speakers 6, 8 or 10 feet apart, those differences would be swamped by room effects. However, sitting right next to each other in a controlled test, those differences will be heard. In my job, we test all the time. Testing has it's own set of methodologies, and one of these is to establish a control. You haven't done it, making the test fundamentally flawed, because you don't know whether the differences you are perceiving existed PRIOR to the "break-in". Sorry, but that's the way it is.

 

[DMF]

Quote:

Originally posted by tonygeno If break-in takes 100's of hours then a 30 minute initial test would not have affected the results.

Well, the test is to find out if there is any such phenomenon as break-in. Your point is valid, but it isn't a fatal flaw. If he does hear a difference, additional testing could establish the cause. If the "raw" speaker is broken in and the difference goes away, I'd call that determinate.


No one is claiming that break-in takes 100s of hours. The 200 hour period was chosen so that in the case no difference is heard, "the pro-break-in crowd" (PBIC?) couldn't argue that the period wasn't long enough. If it takes

 

[rockbottom16]

Quote:

BTW, if anyone thinks that 200 hours isn't enough, speak now or forever hold your piece.

Nope...200hrs is perfect go for it.

I used to mildly believe in break-in for maybe 40hrs or so but this time with my new Kef 9000ACE I kept a time log and sure enough those babies took off after 200hrs

 

[rockbottom16]

Hey Craig,


I was just curious if the type of music played during this test matters. I was playing heavy rap most of the time during the 200hrs since I was breaking in my Velo SPL1200II at the same time.


So maybe break them in with .50 cents might help the process.

Just my .02 cents

 

[tonygeno]

Quote:

Originally posted by craigsub You want facts, here are some facts And quickly, Name the esoteric brand that is matched closer than 0.5 dB...

Vandersteen.


Now, sorry you need to feel so defensive, and no, I don't claim to know everything, but I do know how to set up a valid test. Unfortunately, you haven't.


Let me share a real world experience with you. I received a pair of speakers from a well-known manufacturer. I took the pair out of the boxes and proceeded to calibrate. Strangely though, they sounded quite different from each other. The manufacturer told me to set them next to each other and see if the differences weren't room induced. Well, sitting right next to each other, they still sounded different (drammatically different). The manufacturer, on one of his trips east (he's in Colorado) came over to the house with the speaker's designer (who happens to live 15 miles north of me) and proceeded to disassemble one of the speakers. The crossover had come loose in shipping. Fortunately, they brought a new crossover, swapped it out and we again did a side by side calibration. The differences were again very large. Now we weren't sure if one or both of the speakers wasn't functioning properly. They both went back to the manufacturer since one or both was still not right.


If I had run your test and kept one of the speakers boxed for 200 hours, what would my conclusion have been????: that there was a dramatic difference in sound after break-in.


You see now, why your test is fundamentally flawed? There was no control established at hour 0 of break-in for each speaker. I don't know everything, but I do know testing.

 

[cschang]

I think all Tony is saying is that you need to make sure that both speakers sound the same at 0 hours, at least to your own ears. If they don't, you will not get an accurate assessment.

 

[tonygeno]

Ok, I edited out the Big snip remark. That was gratuitous.

"Regarding YOUR control "experience"... IF you had run the speakers in your example for 200 hours, you would have still had a marked difference between the two speakers, because the difference on your case was a flaw in one or both speakers, not due to break in. IF there is break in, after two hundred hours of both speakers being run together, they should sound the same. Sorry if you don't like that, but it is true."


Yeah, but how do I know that it wasn't break-in at hour 200 for speaker A and hour 0 of speaker B? I'd need to then play B for 200 hours and A not all to then make sure the break-in is indeed real. Using my method, you'd be able to listen in stereo after 200 hours and enjoy the speakers. Using your method, you'll not have final results until 400 hours. If the break-in crowd agrees that it takes hundreds of hours of playing for a speaker to break-in, how would comparing them at 0 hours have ruined your test? At 200 hours you would have had a speaker with 30 minutes on it and another with 200 hours and 30 minutes. If they sounded the same at 30 minutes and different at 200 hours your test would have meant something, as you would have established a control.

 

[tonygeno]

Quote:

Originally posted by cschang I think all Tony is saying is that you need to make sure that both speakers sound the same at 0 hours, at least to your own ears. If they don't, you will not get an accurate assessment.

Thanks, Curtis. That's exactly what I'm trying to say, but I guess I'm not doing a very good job of it.

 

[cschang]

Craig...I understand what you are trying to do.


But....if speaker A and speaker B do not sound the same in the beginning(zero), and they do not sound the same after 15-30 minutes.....or 200 hours, how would you know if break-in had occured?


With every test, ground zero has to be controlled.


Your initial comparing of the two speakers will start at one speaker at 200 hours and the other at 0 hours. The very first comparison should be at 0 hours for each.


Now if there is no difference between 200 and 0 hours...or 200 and 10 hours.....200 and 100 hours....then it is a non-issue.

 

[DMF]

Quote:

Yeah, but how do I know that it wasn't break-in at hour 200 for speaker A and hour 0 of speaker B? I'd need to then play B for 200 hours and A not all to then make sure the break-in is indeed real.

What am I, a potted plant !? Didn't I already say that? Didn't Craig?

Quote:

If the break-in crowd agrees that it takes hundreds of hours of playing for a speaker to break-in

Do they? You know them? Was there a poll I missed?


Using Craig's method, we also control for the possibility that break-in for these speakers is very quick. And believe me, he has more then enough speakers to listen to while the raw one catches up, should that prove an issue.

 

[Bigus]

craig,


You can safely ignore tony's complaint, as your proposed test method does offer that control. As you have pointed out, you only need to verify that the speakers are the same either before [/b]or[/b] after they have both played for the same length of time. It doesn't really matter which.


The only disadvantage of comparing after both are broken in is that if the still sound different, the test really doesn't offer any conclusive results one way or the other. If they continue to sound the same, or if one is different and then in time sounds the same, then the test produces meaningful results. If they remain different, all you really know is that the speaker matching was poor and that they may or may not have drifted relative to one another during testing.



A point I'd like to stress, however, is that it isn't sufficient to simply note "same or different" for each trial or session. After all, a bias that leads you to believe break-in is real might have you thinking "different" at first and "same" later on.


You need to be able to offer proof that your ability to distinguish a difference is real, not imagined. The only way to do so, in the context of your test, is to consistently identify the two speakers absolutely. In other words, if your very first listening session leads you to believe they are "different" then you need to try and label them A and B. The next session, you need to try and match your A and B labels. That kind of stinks, but that's how it works. You need to try and focus on some aspect of the sound that you believe makes them different (if you think you hear a difference). Note which speaker belongs on which end of the spectrum for whatever aspect you choose (for example, A is brighter, B less so, and in each trial you try to consistently label the bright one as A).


With three speakers you could arrange an ABX test that would be much easier on you, at least, but with two you need to have some auditory memory from one trial to the next.


I hope you understand my point. You can't simply record "same or different" each time, because technically they will always be different, and there will be nothing objective to match your results to. You have to record "this speaker sounds X, so it is A" each time, and hope for consistency.


Of course, with the dramatic break-in claims made, you would think that it will be easy to hone in one one aspect of the sound and use it as an identifier to label A and B correctly. You might be surprised.

 

[DMF]

Quote:

if the still sound different, the test really doesn't offer any conclusive results one way or the other.

I disagree there. One would have to conclude that the test was invalid.



Good point on identifying each one, though.