 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
69.181.86.110
In Reply to: Re: Nope posted by Jim Austin on October 22, 2005 at 09:15:41:
Thanks for clearing up your meaning. Double blind actually has a different meaning in the wine world, but we do everything weird. Your definition is the one I will usehere.The sort of analytical tasting you ask about is done routinely. For example, someone wishing to judge in several of the major competitions must be able to demonstrate in a blind tasting that they can detect and identify threshold elements like volatile acidity, TCA taint, brettanomyces, and the like via spiked samples at known concentrations. Yes, ABX is used, but it's one of several analytical tasting formats. Duo-trio and triangle are even more common.
In my own work (wine packaging), we are specifically looking for differences in most of our tests, not just better/worse. Our sensory panel consists of people who, under double blind conditions, are reliably able to distinguish the differences we're testing for. And any test that we or our customers would take seriously must be run double blind.
Follow Ups:
> > someone wishing to judge in several of the major competitions must be able to demonstrate in a blind tasting that they can detect and identify threshold elements like volatile acidity, TCA taint, brettanomyces, and the like via spiked samples at known concentrations. < <...the wine folks are ahead of us audio folks in some respects, maybe because in the audio world there are no formal competitions. And though some wine mags may administer tests to prospective reviewers (do they?) I doubt it's very common for folks in the press to be credentialed.
I've got to say that it probably wouldn't be that hard to develop a pretty basic set of "elements" for audio reviewers--jitter levels, broad-band (but low-level) midrange enhancement/depression, rolled off treble, boomy bass...whatever. I suggest that with trepidation, but I suppose that as long as we get to practice first I'd be okay.
What's most interesting about the idea is that none of those things the wine folks are tested for by themselves guarantee that the judge can tell a good wine from a bad one. But they DO show that the pallette is sensitive to the "colorations" common to wine, which is a pretty good start. Beyond that it's all subjective.
One more point, so that folks don't think I'm changing my mind on this. What doesn't work in audio--and I suspect would not work with wine either--is routine, statistically rigorous blind testing/tasting as a means of routinely evaluating wines. There's certainly an important role for blind testing in audio RESEARCH, in trying to determine what really matters for good sound. And there may be a role in it (though this hasn' happened yet) for "credentialing" potential reviewers (and other industry folks).
Just the other day I tried to convince a cable manufacturer I was visiting with to do some blind cable tests in their amazing listening room. The guy I was talking to (not one of the principles but a fairly high-level employee) claimed that you could hear the difference between a decent generic (monster wire or similar) and one of their mid-range twisted-pair constructions "with a boombox." Assuming that's true, there should be no problem proving to folks that cables really DO make a difference. If you can hear the difference with a boombox, surely you can pass a blind test with good gear. They declined.
Maybe my views HAVE changed a little. I'm impressed that the wine folks do so much blind testing, even some of the quantitative kind.Best,
Jim
Another method we sometimes use for keeping tasters honest is to throw the same wine into the mix twice. I once failed this spectacularly, giving a wine I had blasted as rubbish the day before a gold medal.Not as bad as a former employer of mine who once gave an absolutely damning crit. of a wine at a public exhibition only to find it was one of hers.
Mark Kelly
Given the varied opinions of the same audio system among "experts," I'd have to wonder.'Course the same thing may apply in wine tasting, for all I know.
"What doesn't work in audio--and I suspect would not work with wine either--is routine, statistically rigorous blind testing/tasting as a means of routinely evaluating wines."Actually it does work in routinely evaluating wine. Most reviewers do (or claim to do) their tastings for publication blind. Same thing with the analytical stuff that we do at my company.
DBT means "double blind test", no more, no less.Some people equate ABX testing with DBT's.
A lot of people falsely equate ABX testing with sequential ABX testing.
Some people insist that, despite the tons of evidence regarding the length of primary auditory memory (200 milliseconds give or take), "fast switching hides results" which of course is as far from what the serious types have figured out as you can get, but so it goes.
And then we have the people who talk about how it stresses them.
A real DBT, done right, will be double blind, it will allow seamless switching, under the control of the subject, and will allow the subject as little or much time as they want.
And it doesn't have to be ABX.It can be ABC/HR, it can be AB-signal detection, there are more ways to skin a cat than most of them here imagine.
But "double blind test" means, simply, that neither the test administrator that's present during the test (if any) and the subject know what the probe stimulii is. They may, and should, know what all reference stimulii are at all times.
Now, ABC/hr, ABX, and AB-signal detection tests (for probes that apply) have all been shown to work with noise, tones, music, single instruments playing single notes, complex soundfields with music and speedh (and that as far back as the 1920's), and pretty much any sort of acoustic stimulus.
It's only the high end that denies the evidence of the matter.
Btw, I like DeLille, perhaps it's a question of taste. What sort of cab/zin/bordeaux-like red do you like?
> What sort of cab/zin/bordeaux-like red do you like?>Zins rule! They don't need food to be enjoyable. There's nothing like a nice glass of zin to go along with your music on these cool autumn evenings. My personal favorites are the Sonoma Dry Creek Valley Zins, like Rafanelli, and a number of the Rosenblums from the Bay Area, like the Harris Kratka and Rockpile vinyards.
Anyone else have any favorites?
But the last few bottles I've found from newer years weren't so hot.
Don't miss Mick Unti's lovely zins.
You anywhere around Boston?clark
Nope. Napa. But I do make it there from time to time.
These were all points inherent to my argument. Double blind is universal, but the format of the test will vary, depending on what the experimenter is trying to probe.As far as a Bordeaux blend... how about, ummm, Bordeaux? I've becaome less enamored of Bdx in recent years as the region has prostituted itself to the points gurus, but there are still rays of light here and there. When I drink one of the old California Cabernets, I shed a tear realizing that no-one is doing it like that anymore. Nowadays, it's extract, oak, and alcohol, plus a bit of residual sugar. Bigger is better. Balanced, nuanced wines are a thing of the past.
Excuse me, I hear a '77 Carneros Creek Fay Vineyard cabernet calling my name...
I do hear what you say, indeed.My old Simi Cabs had to go (well, in a very nice way) when I left where I used to live... But I miss them.
I'm not a big fan of the Big Swinging Dick appellations, nor do I drink much Cab Sauv these days. My focus is more Northern Rhone, Loire Valley, and Italian country wines for reds, Germany, Loire, and Austria for whites. The secret is to look at the so-called second tier of appellations- St-Joseph, Cornas, Crozes. And the vins de Pays. Look for St-Joes from Faurie, Cheze, Gaillard, Monteillet, Pichon, and Tunnel (Stephane Robert), Crozes from Graillot. Ogier's La Rosine is an astonishing Vin de Pays. Other hint: 2001. Most critics missed it, but the Rhone winemakers think it's the best vintage since '78 (at least).In the Loire, just buy anything you see with a Louis/Dressner import label on it, red or white.
> In my own work (wine packaging), we are specifically looking for differences in most of our tests, not just better/worse. Our sensory panel consists of people who, under double blind conditions, are reliably able to distinguish the differences we're testing for. And any test that we or our customers would take seriously must be run double blind.>...are you familiar with the types of testing done in the food industry which sounds similar to what you are involved in? Much of it is blind, but there are numerous other protocols used far removed from an actual DBT. And some of these may have more relevance to audio than the DBT, as well. A blind A-B-A-B test, for example, would be a blind, serial monogomous test, IIRC. As described above, an audio DBT is most often an ABX-type test, where the tester must match X with A or B.
There are indeed lots of formats possible as I mentioned in the post above. The choice of best test format depends on what question you're trying to answer. Defining that clearly right at the beginning is key.
This post is made possible by the generous support of people like you and our sponsors:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
