[daj]

This is meant as an effort to gauge how people feel about the applicability of DBT to determining whether individual audio effects are real, such as whether adding an extra power line filter to a particular preamplifier under a particular set of noise conditions has an audible effect.

This is just a poll and not meant as an expression of opinion, nor as an invitation to discuss DBT, nor to discuss the merits of discussing it.

If you are interested in knowing how this group feels about it on average, then please participate in the poll.

I noticed Occam's abolition of DBT discussion after my initial posting of this poll. Being a poll, it is free of expressions of opinion; on the contrary, it's for learning about opinion. I take Occam's point well about avoiding skirmishes and unpleasantness, and vow not to discuss DBT here, nor to engage participants who feel one way or another about it. Occam has decided to allow this poll to stand for now, so there's no need to jump in on his behalf.

Background:

Double blind testing is a way of comparing two conditions that seeks to eliminate spurious influences. Neither the experimenter nor the subject knows which condition is which until after the testing is over. In the case of audio effects, where one is unlikely to retain a memory of each condition for more than a few seconds, the conditions are presented very close in time and repeated many times. Also, the levels and content of the two conditions are matched very closely. Some qualification of the subjects can be useful, such as picking sets of professional recording engineers, musicians, and other specific groups, as well as general audio hobbyists.

The simplest interpretation of the data is that when the subjects can determine more than half the time which is which, there is likely to have been a real difference. More precisely, how likely it is that a difference was heard is a function of the number of trials, the number of subjects, and the percentage of correct identifications. Such a test is most easily conducted under computer control, where the computer picks the sequence of conditions at random.

The most basic testing as described above does not aim to determine how intense an effect is, only how likely is its existence. Trials can, however, include ratings of intensity by the subjects at a cost of greater complexity in data reduction, where each intensity rating acquires a likelihood of existence.

The results of this poll will be available after 30 days. Thanks for participating.

The poll was modified about three hours after it started and the votes reset to zero.

Sorry about the ambiguous title, which could have been more clear but a lot longer, such as: Separating Audio Realities that Inspire Romantic Experience from Romantic Notions of Audio Reality

[ohenry]

http://www.audiocircle.com/index.php?topic=14093.0

[daj]

http://www.audiocircle.com/index.php?topic=14093.0

Okay, thanks, ohenry. Added a note about that. It's just a poll, though. I'm not grandstanding. I only want to know what people here think.

[Scott F.]

daj,

Not to steal someone else's quote but there are a few absolutes in life

1. Death
2. Taxes
3. Any posts on cables, power cords or DBX testing will get locked or binned

As Occam has mentioned in the rules, these discussions are absolutely futile as they only end in a flame war.



Be prepared if this gets deleted.....



edit;   I replaced 'to be' with 'if this'

[daj]

daj,

Not to steal someone else's quote but there are a few absolutes in life

1. Death
2. Taxes
3. Any posts on cables, power cords or DBX testing will get locked or binned

As Occam has mentioned in the rules, these discussions are absolutely futile as they only end in a flame war.



Prepared to be deleted.....

Well, Scott, I appreciate that, having recently put some such flames behind me. In this case, there's nothing up for discussion, since it's only a poll. I just want to learn what the opinions of others are with a minimum of fuss and nuance. In fact, the poll will be more effective without a backdrop of opinion, so it is best if this does not turn into a discussion. I hope for your understanding, and I hope at least some others are as interested in the results as I am.

[Wind Chaser]

Question: Are double blind tests required to determine the existence of an audio effect?

Hmm, I don't know how to answer that but I'm inclined to say no from the perspective of how I come to understand the differences between different components.  Getting to know how a component sounds takes time, much like getting to know a person.  The more time you spend with someone / something, the better your knowledge.

[daj]

Question: Are double blind tests required to determine the existence of an audio effect?

Hmm, I don't know how to answer that but I'm inclined to say no from the perspective of how I come to understand the differences between different components.  Getting to know how a component sounds takes time, much like getting to know a person.  The more time you spend with someone / something, the better your knowledge.

Sorry, it does sort of sound like I might have been asking about applicability to evaluating and understanding the differences between components. It's only to detect whether there is any difference.

For clarity I should have mentioned that one would be testing for whether there is any effect from a single specific change, like two different AAC compression bit rates, two different pairs of speakers, two different preamplifiers, etc. Will add that into the description, now.

[Occam]

Dave,

I'm going to let the poll stand for the time being. I'd be far more trusting of your motives if I hadn't viewed your posts, which are, IMO, patronizing and add naught to the discussions.

I'll also add the following - Those who invoke the DBT shibboleth rarely have any actual experience in setting up and conducting those tests in any way that could be construed as valid. Personally, I'm a great believer in ABx testing, and its straightforward to perform such for (pre)power supply filtering design and attempt to correlate the results with actual measurements. Other, audio specific, areas involving subjective testing are far more difficult and largely beyond my ken.

And how would invoking democratic principals, a poll, add anything? For marketing, yes, but technology??? Maybe I should start a poll  'Extended Beta in BJT EFs, Yea or Nay?'
Why would I give a farthing for the results of that poll? or this poll?
Howzabout a poll on what everyone thinks Planck's Constant should be?

And Yes, I'll probably move this thread to the Fight Club as some point, as it inevitably sprials out of control. Hopefully, the poll can continue in that circle.

[daj]

Dave,

I'm going to let the poll stand for the time being. I'd be far more trusting of your motives if I hadn't viewed your posts, which are, IMO, patronizing and add naught to the discussions.

I'll also add the following - Those who invoke the DBT shibboleth rarely have any actual experience in setting up and conducting those tests in any way that could be construed as valid. Personally, I'm a great believer in ABx testing, and its straightforward to perform such for (pre)power supply filtering design and attempt to correlate the results with actual measurements. Other, audio specific, areas involving subjective testing are far more difficult and largely beyond my ken.

And how would invoking democratic principals, a poll, add anything? For marketing, yes, but technology??? Maybe I should start a poll  'Extended Beta in BJT EFs, Yea or Nay?'
Why would I give a farthing for the results of that poll? or this poll?
Howzabout a poll on what everyone thinks Planck's Constant should be?

And Yes, I'll probably move this thread to the Fight Club as some point, as it inevitably sprials out of control. Hopefully, the poll can continue in that circle.

Okay, thanks for bearing with me for the moment, Occam.

About my less than helpful previous posts, I see my mistake, and those posts and their attitude are in the past. I thought my long one about absolute polarity was pretty good, though. http://www.audiocircle.com/index.php?topic=37363.msg337704#msg337704

About my motives, I am attempting to grock the the current audio zeitgeist. 

About this being a poll on arcane technology, I think it's actually a poll about whether scientific certainty is important in audio, with the somewhat unfortunate underlying assumption that DBT is at least potentially a way to attain this certainty.

Your incisive point makes me realize that I probably should have asked that question directly instead of invoking the DBT shibboleth, as you so aptly put it. Didn't even think of it, sigh. Rather than make the substitution, now, though, I'd like to let the current poll ride and then compare the results to those of a later, explicit poll on scientific certainty, since as a side effect, this may provide an interesting, additional clue as to whether DBT is regarded as being a route to scientific certainty.

About how straightforward ABx is for the whole range of audio effects, I don't think it would further bias the results of this poll to say that any effect can be gauged straighforwardly once one has developed a suitable apparatus for swapping the conditions without human awareness and without introducing additional effects, and I can imagine such an apparatus for every audio aspect I know of.

Razor on.

[fredgarvin]

As for the title of this thread, Romance and reality are eternally joined in varying degrees in the shrinking ranks of audiophiles. Love of gear and of music are intrinsically fueled by the romantic and by the quest for the unobtainable. Objectivity faces a struggle there.

[shep]

Without the romance the reality is pedestrian, utilitarian and boring. I just cannot get the mindset that wants to take the magic and the "mystery quest" out of our hobby. Every time this relentless issue comes up I have the same reaction.

[Daygloworange]

daj,

I found your post on "Absolute Polarity" really interesting. It's given me a lot of pause for thought.

The one on the Bolder cable circle, I could have done without, however....

I'd like to see you post more often along the lines of the former. Good stuff.

Cheers

[daj]

daj,

I found your post on "Absolute Polarity" really interesting. It's given me a lot of pause for thought.

The one on the Bolder cable circle, I could have done without, however....

I'd like to see you post more often along the lines of the former. Good stuff.

Cheers

Thanks. It was fun to think about, too.

[Kevin Haskins]

I think DBT is a useful tool but just like any single experiment, it isn't conclusive on its own.    You need a lot bigger sample than one testing scenario to get a meaningful picture.   Its biggest use is in research, not in consumer audio choices.   

It is helpful to remove your own bias though.   I don't hold rigorous standards but I like to use single blind testing as a quick check about my preconceived notions when designing something.   When I make a change that I think is an improvement my bias is to hear an improvement.   Doing some quick SBT allows me to see if my perception is good enough to verify the change.   Its very difficult to do a scientifically valid test though and it would quickly become cost prohibitive if I had to hold myself to that standard on every design chore. 

[daj]

I think DBT is a useful tool but just like any single experiment, it isn't conclusive on its own.    You need a lot bigger sample than one testing scenario to get a meaningful picture.   Its biggest use is in research, not in consumer audio choices.   

It is helpful to remove your own bias though.   I don't hold rigorous standards but I like to use single blind testing as a quick check about my preconceived notions when designing something.   When I make a change that I think is an improvement my bias is to hear an improvement.   Doing some quick SBT allows me to see if my perception is good enough to verify the change.   Its very difficult to do a scientifically valid test though and it would quickly become cost prohibitive if I had to hold myself to that standard on every design chore. 

Oops, thanks, Kevin, but a little too much going on, there. Trying not to have a discussion of blind testing, here, which would go against the rules of the Lab (not ironically, but to avoid confrontation), and also to keep the progress of the poll itself blind until it's done.

[daj]

Without the romance the reality is pedestrian, utilitarian and boring. I just cannot get the mindset that wants to take the magic and the "mystery quest" out of our hobby. Every time this relentless issue comes up I have the same reaction.

I'm sorry about that title, but it was a convenient rhetorical device, and precise wording would have made the title far too long.

Of course, reality and the romance it inspires are inseparable. The idea behind the title is that on the one hand, there are aspects of audio that hold up to a stringent test of reality and contribute to the romance of the experience of music and sound. On the other hand, there are aspects that are taken as reality that may not hold up to a stringent test and are thus themselves only romantic notions. Not to denigrate these latter notions, the question is only whether scientific method is required to know which is which.

The title could have been: Separating Audio Realities that Inspire Romantic Experience from Romantic Notions of Audio Reality

[Kevin Haskins]

Oops, thanks, Kevin, but a little too much going on, there. Trying not to have a discussion of blind testing, here, which would go against the rules of the Lab (not ironically, but to avoid confrontation), and also to keep the progress of the poll itself blind until it's done.

Sorry... how about this.

As a consumer, No... all that is valid is your own personal experience.

As a designer, Yes.. you need to know what is real and what is imposed by your own pre-conceived notions.

As a marketing guy, No... all that matters is what the customer believes.  Physical reality is irrelevant and sometimes harmful to your bottom line.

So the entire debate is relative to where you stand. 

[daj]

Oops, thanks, Kevin, but a little too much going on, there. Trying not to have a discussion of blind testing, here, which would go against the rules of the Lab (not ironically, but to avoid confrontation), and also to keep the progress of the poll itself blind until it's done.

Sorry... how about this.

As a consumer, No... all that is valid is your own personal experience.

As a designer, Yes.. you need to know what is real and what is imposed by your own pre-conceived notions.

As a marketing guy, No... all that matters is what the customer believes.  Physical reality is irrelevant and sometimes harmful to your bottom line.

So the entire debate is relative to where you stand. 

Okay, interesting insights, Kevin. I guess the takeaway is that, like any poll with only one question, no frame of reference, and no conditions, this one is too simplistic to convey nuance. I trust that most will vote from their own POV, whatever that is, and we'll just have to make what we can of the results.

[Scott F.]

daj,

Since Occam has allowed the post and poll to stand, I thought I'd go ahead and participate 

I voted Yes, when the effect is subtle. I have more equipment and accessories run through my house each year than most people. Most often, the differences are pretty plainly audible. There are times though when it's really tough to distinguish if there is a change. I wonder to myself if I'm imagining things or if I actually hear something. When it gets to that level of subtlety, I usually call things even (no change) even though I might perceive a slight difference.

I do use a couple of devices that aren't on the DBX'ers approved list but it helps me distinguish (if there is a difference) between items. I use a NEC AVX 910 source switcher that has a remote. I do my best to level match spl's and then I start flipping back and forth. Not perfect and won't pass anyones true scientific standards but it's good enough for me to pick up some of those slight differences in presentation, tone, detail or whatever.

The other item I use is a simple Adcom or older Niles speaker switching unit (on rare occasions) The Niles has level pots and is transformer based. This is a bit tougher to use since level matching can be a chore if not impossible. After using it, and taking into account the sensitivity and impedance differences between speakers (preset volume levels on the pre or the Niles) I can come close to the same levels. Again, it gives me an idea of the differences in some speakers.

Ultimately, none of this really matters as our own preferences in sound outweigh any change we may perceive. Things like cottage industry caps and resistors change the sound. Change does not necessarily equate to better. The real questions are, do you like it and if so is it worth the money?

So goes audio, there are no absolutes, only opinions based on our own biases.


Hope I didn't step too far into the whole DBX netherworld with that reply 



Oh daj, something you probably aren't aware of here at AC. The Council requires participating members who are manufacturers to list their business name and a link to their website in their signature line. Take a peek at Kevin's signature line and avatar and that will give you an idea what they are looking for. You can change that in your Profile. See the tab at the top of the page that says.... Profile 

[Kevin Haskins]

Oops, thanks, Kevin, but a little too much going on, there. Trying not to have a discussion of blind testing, here, which would go against the rules of the Lab (not ironically, but to avoid confrontation), and also to keep the progress of the poll itself blind until it's done.

Sorry... how about this.

As a consumer, No... all that is valid is your own personal experience.

As a designer, Yes.. you need to know what is real and what is imposed by your own pre-conceived notions.

As a marketing guy, No... all that matters is what the customer believes.  Physical reality is irrelevant and sometimes harmful to your bottom line.

So the entire debate is relative to where you stand. 

Okay, interesting insights, Kevin. I guess the takeaway is that, like any poll with only one question, no frame of reference, and no conditions, this one is too simplistic to convey nuance. I trust that most will vote from their own POV, whatever that is, and we'll just have to make what we can of the results.

I'm being a stinker.   You can ignore me.... just feeling a little ornery today.