[Joe Bloggs]

You can tell the greatness of a man by what makes him angry.

-Abraham Lincoln​

--A quote a head-fier from AAMML puts in his sig. Thought it's a nice quote that could give some of us here some perspective on what we need and need not do with head-fi.

----
Another year, another skeptics lounge thread locked?

I think the rationale behind the "lounge" may not be entirely healthy--it seems to be a place where the gang gathers after a day at the trenches, knocking down beers in chain-fire fashion and grumbling about the bad day they had. (except you can't grumble too hard, because, well...)

When we were not grumbling in the thread, we were hatching crazy ideas to "take down" the "enemy" (bigshot's (RIP) likebomb comes to mind), which is of course also frowned upon by the mods.

If this is to become the next "general talk" thread of Sound Science, I would like to see neither of the above here.

I think the only allowable response to posts of "egregious offenses on good audio common sense" on head-fi (or whatever you would like to call those posts against which you'd like to rally a call to arms, or something) is to respond evenly with good arguments. Any name-calling or similar sub-argumentative tactics is never allowed, whatever the other side does.

But if you think about it, calling names has never won anybody an argument or drawn many people to their side. On the other hand, one could possibly win over some people with truly well-phrased arguments. Or, if the "opposition" is not truly interested in the argument (as may be the case with a true "troll"), they may instead be bored into submission / posting elsewhere.

Going by the numbers, and the fact that any member of head-fi can post anywhere they dang well please, the fact that this place is called "Sound Science" obviously doesn't mean we own the joint, only that we are allowed to express our opinions relatively freely here.

The thread title was written to reflect the more "intelligent" nature of the discussion I'd like to see here.

If you'd like a good old lounge to grumble and knock down beers in, I suggest going elsewhere.

 

[Steve Eddy]

Just joined. Can you capitalize the f in HeadFi and remove the hyphen? Thanks!

EDIT: Oh, wait, leave the hyphen. I see that it's official. Just capitalize the f.

se

 

[headdict]

[Mod Edit: Read the Posting Guidelines please.]

 

[Joe Bloggs]

Even if they may never agree with you, you may still earn their respect...

http://www.head-fi.org/t/486598/testing-audiophile-claims-and-myths/6495#post_11644737

 

[Steve Eddy]

Even if they may never agree with you, you may still earn their respect...

He's been on the receiving end of some pretty nasty stuff here. I'm not at all surprised he gave that post a thumbs up.

I hadn't really read much of any of his posts, but saw the Schiitstorm that surrounded them here. And I was just going to write him off as a troll as well. But then I read a couple of his posts that he was getting slammed for. Had to do with cables foe electrostatic headphones. I started to pile on and then realized he was absolutely correct.

That caused me to want to get to know more about him. So I PMed him, and shortly thereafter ended up having a long telephone conversation with him.

He's a genuinely nice guy. And in terms of general audio, actually quite knowledgeable. After our telephone chat I considered him a friend.

He has his faults (who doesn't?). Namely that he puts too much faith in human subjective perception and his knowledge of materials science comes from audiophiles. But I'm working on him.

All I can say is that I'm glad I stopped and took the time to get to know him better. Sadly, some still see fit to call him a troll.

se

 

[RRod]

I think we should discuss the fine points of suggesting ABX tests to wanderers-in. I've been reading about some ABX-related fusterclucks over at hydrogenaud.io: the keys jangling test, the AIX test, and an AES paper with some iffy statistics (see here for starters). They hit on many of the roadblocks to properly using the protocol, but the first two especially highlight the perils of mixing what is supposed to be a rigorous test and the Interwebz. Part of me feels like we need a sticky to point to and say "here is the checklist for how to do an ABX test of X vs Y by yourself," with the added caveat that the rest of us can't really take a person's results as proof of anything, given how easy these things are to do improperly or to rig. Thoughts?

 

[Joe Bloggs]

I think we should discuss the fine points of suggesting ABX tests to wanderers-in. I've been reading about some ABX-related fusterclucks over at hydrogenaud.io: the keys jangling test, the AIX test, and an AES paper with some iffy statistics (see here for starters). They hit on many of the roadblocks to properly using the protocol, but the first two especially highlight the perils of mixing what is supposed to be a rigorous test and the Interwebz. Part of me feels like we need a sticky to point to and say "here is the checklist for how to do an ABX test of X vs Y by yourself," with the added caveat that the rest of us can't really take a person's results as proof of anything, given how easy these things are to do improperly or to rig. Thoughts?

Aside from music file tests....................................................................................

I've never even bothered to do any ABX tests for myself, I knew before anybody told me that it would be quite impossible for me to conduct any proper ABX tests of anything of a mechanical nature. Asking a believer in differences in your chosen topic of known indifference (excuse me for having to tread carefully ) is just an invitation for him to run a non-level-matched, improperly blinded test, come back with a nice 9/10, say "ha! There's your proof!" and the discussion to go further downhill while he becomes firmer in his belief than ever.

We obtained our knowledge of the non-difference of most components through a combination of simple electronics theory and well-documented, properly run ABX tests already conducted on large batches of test subjects, published on the AES journal and the like. Probably better to point to those in most cases than to ask the impossible of someone who's probably all thumbs when it comes to experiment protocol.

 

[RRod]

I think having a primer on self-ABXing music files would be useful, since that's the most typical case (on here at least) where people might actually go through the necessary steps. Hardware is of course a whole different beast, but even there having some information around for the willing can't hurt them, at least.
I guess I can't envision people NOT suggesting ABX at times on this forum, so I thought a bit of discussion about how such suggestions should go might be useful.
 
As far as paper referencing, it  would see that if we link to this:
http://www.aes.org/e-lib/browse.cfm?elib=14195
a cheeky subjectivist can link to this:
http://www.aes.org/e-lib/browse.cfm?elib=15398
then we're off pointing out the procedural and statistical superiority of our favorite AES paper. Perhaps that is unavoidable.

 

[Steve Eddy]

But let's not get too hung up on ABX.

I think difference testing should be the go to first choice. If the residual isn't audible, there ipso facto can't be an audible difference. If there is an audible residual, then that would warrant moving on to ABX testing to see if the difference is audible while listening to music.

se

 

[RRod]

But let's not get too hung up on ABX.

I think difference testing should be the go to first choice. If the residual isn't audible, there ipso facto can't be an audible difference. If there is an audible residual, then that would warrant moving on to ABX testing to see if the difference is audible while listening to music.

se

 
 
I wouldn't know on the hardware end how much easier a null test is to set up compared to an ABX. For music, I worry a bit about the double-testing aspect, at least from a sociological view. If someone nulls something like an mp3, they will certainly hear obvious differences, which means they will have to want to believe us that an ABX might still show these differences to be inaudible in the context of the music. Sending them straight to an ABX in something like Foobar might be better in such a case. But you're right that the silence of a difference file is a powerful result, and for testing things like pre-ringing from Redbook conversion, it is probably the way to go.

 

[limpidglitch]

  I think having a primer on self-ABXing music files would be useful, since that's the most typical case (on here at least) where people might actually go through the necessary steps. Hardware is of course a whole different beast, but even there having some information around for the willing can't hurt them, at least.
I guess I can't envision people NOT suggesting ABX at times on this forum, so I thought a bit of discussion about how such suggestions should go might be useful.
 
As far as paper referencing, it  would see that if we link to this:
http://www.aes.org/e-lib/browse.cfm?elib=14195
a cheeky subjectivist can link to this:
http://www.aes.org/e-lib/browse.cfm?elib=15398
then we're off pointing out the procedural and statistical superiority of our favorite AES paper. Perhaps that is unavoidable.

 
I was about to ask if that 44.1/88.2 test was what you were referring to.
Do you know any more about the statistics than what was discussed in the HA thread? It looks a bit of a mess.

I think you're absolutely right about ABX tests. They are a great tool, when used correctly, but we have no way to assess the validity of any proof presented to us, so it makes no sense for us to ask for it.

 

[maverickronin]

I don't think it will matter what kind of test you propose.  Someone will just make up an ad hoc excuse to explain away results they don't like.

 

[StanD]

  I don't think it will matter what kind of test you propose.  Someone will just make up an ad hoc excuse to explain away results they don't like.

Yes, I wouldn't put anything past some of the true believers of myths.

 

[RRod]

  I don't think it will matter what kind of test you propose.  Someone will just make up an ad hoc excuse to explain away results they don't like.

 
I guess it's the trade-off between being actually helpful to those who want to learn versus putting up with the people who want to make the excuses. If even one person saves some money on a DAC or amp because of a thread, then that's great, even if someone else is making up their test results.
 
 
Quote:

   
I was about to ask if that 44.1/88.2 test was what you were referring to.
Do you know any more about the statistics than what was discussed in the HA thread? It looks a bit of a mess.

I think you're absolutely right about ABX tests. They are a great tool, when used correctly, but we have no way to assess the validity of any proof presented to us, so it makes no sense for us to ask for it.

 
Haven't seen the paper yet, so I can't expand on the thread comments. But yeah, seems a mess in there, even beyond the statistics. Any time a result is "well it was significant if we go 2-sided because the only people who passed got scores that were too LOW", I get a bit worried.

 
Recently someone showed me a pass of an ABX of 320lame versus hi-res on a choral music track, where they claim they heard the differences in the reverb via extremely quick switching. Do I believe him? Do I flood him with questions on possible issues? In the end the easiest thing to do it just let him have his results and be happy, and just continue on knowing that I myself couldn't hear anything. Still, any information either of us got about properly executing the test was useful.

 

[maverickronin]

Possibly on topic, what do my fellow skeptics see as their purpose in engaging people on forums like these which are so overrun with nonsense?
 
One is very unlikely to convince anyone to give up their irrational beliefs in one conversation whether it's online or in RL so many see it as pointless.  It usually goes like this.
 
 

 
 
OTOH, even if you don't convince the person you're arguing with there's another factor which is commonly overlooked, the audience.  Lurkers are watching in the background.  They may be lightly committed to one position or another or may be completely new to field and not yet know left from right.  Those are the people who are most likely to be swayed in one way or another.  When I was more active on here a few years ago I used to get PMs from people I hadn't previously interacted with in threads thanking me for talking straight, saving them from buying snake oil, and asking technical questions so I can attest that does work.
 
On the other hand, doing that kind of thing can pretty exhausting and I seem to be getting more and more jaded and cynical about the prospects since it's hard to keep up that kind of effort.
 
What does everyone else think?