[Bullseye]

What does DBT stand for?

 

[PhilS]

Quote:

Originally Posted by Bullseye /img/forum/go_quote.gif What does DBT stand for?

"Double blind test."

 

[nick_charles]

Quote:

Originally Posted by Bullseye /img/forum/go_quote.gif What does DBT stand for?

A DBT (Double Blind test) is a test where neither the experimenter nor the experimental subject know the identity of the object under test. In medical trials for instance some subjects get a treatment and some get a placebo but the administrator does not know which is given to which patient.

 

[Pio2001]

Quote:

Originally Posted by nick_charles /img/forum/go_quote.gif Ivor Tiefenbrun please note !

My mistake : it was not an Ultra-curve pro, but a Behringer DCS2496.

(AN converter + equalisation -12dB, Q=4 at 40, 100, 1k, 3k, 10k + equalisation +12dB, Q=4 at 40, 100, 1k, 3k, 10k + NA converter) versus cable.

The gain of the DSC had to be set to +0.9 dB for the levels to be matched.

Quote:

Originally Posted by nick_charles /img/forum/go_quote.gif A DBT (Double Blind test) is a test where neither the experimenter nor the experimental subject know the identity of the object under test.

More precisely, a double blind test is a test where there can't be uncontrolled influences between the subject and the identity of the stimulus tested.
Having the experimenter ignore it himself is a good way to acheive the double blind condition.

 

[analogsonar]

I have two storages full of cheap receivers and integrated amps from the likes of onkyo, denon, carver, marantz, cambridge, sony, technics. each one, when swapped out in the same system, has a very distinct difference in sound, noticeable right away when changed out. i have a pair of speakers that took me 8 months to find a receiver that would sound good driving them, because my warm amps made them sound like they had a thick blanket over them. i finally found a carver that worked perfectly.

 

[PhilS]

Quote:

Originally Posted by Pio2001 /img/forum/go_quote.gif Having the experimenter ignore it himself is a good way to acheive the double blind condition.

Isn't the theoretical problem that the experimenter may not be able to ignore it and my unintentionally communicate something to the subject? Isn't that why it is necessary that the experimenter be "blind" also?

 

[jcx]

exactly - called the "clever Hans effect" Clever Hans - Wikipedia, the free encyclopedia

humans are better than horses at interpreting clues about other people's expectations (at least human females)

so the only way to avoid the problem is to arrange the test such that no one in the testing room, test administrator or subjects, has any clue to which instance of the test variable is presented during each trial

afterwords the test scores are comapared with the master record of the test sequence

even automated ABX tester setups need careful vetting to determine if there are clues such as switching delay, relay cliks or noise level differences that allow descrimination of A from B channels independent of the actual tested variable

 

[nick_charles]

..

 

[Pio2001]

Quote:

Originally Posted by PhilS /img/forum/go_quote.gif Isn't the theoretical problem that the experimenter may not be able to ignore it and my unintentionally communicate something to the subject? Isn't that why it is necessary that the experimenter be "blind" also?

The other solution is to have the experimenter in another room.

Quote:

Originally Posted by jcx /img/forum/go_quote.gif even automated ABX tester setups need careful vetting to determine if there are clues such as switching delay, relay cliks or noise level differences that allow descrimination of A from B channels independent of the actual tested variable

True. In one of Matrix-hifi's meetings, an ABX test succeeded between to CD players. They discovered that the line inputs of the Mark Levinson preamplifier in which they were plugged had a built-in fade-in effect, and that one of them was faster than the other.
They redid the test on two identical line inputs and it failed.

 

[milkweg]

Quote:

Originally Posted by jcx /img/forum/go_quote.gif exactly - called the "clever Hans effect" Clever Hans - Wikipedia, the free encyclopedia humans are better than horses at interpreting clues about other people's expectations (at least human females) so the only way to avoid the problem is to arrange the test such that no one in the testing room, test administrator or subjects, has any clue to which instance of the test variable is presented during each trial afterwords the test scores are comapared with the master record of the test sequence even automated ABX tester setups need careful vetting to determine if there are clues such as switching delay, relay cliks or noise level differences that allow descrimination of A from B channels independent of the actual tested variable

If it is required to be that meticulous then I feel pretty safe in saying that if there is any difference then it is so small that it is not worth spending big bucks for. If I was to spend 4G on an amp and already had an amp that cost 1G then I would expect a BIG difference for the 3G difference and not a minor one.

 

[PhilS]

Quote:

Originally Posted by milkweg /img/forum/go_quote.gif If it is required to be that meticulous then I feel pretty safe in saying that if there is any difference then it is so small that it is not worth spending big bucks for. If I was to spend 4G on an amp and already had an amp that cost 1G then I would expect a BIG difference for the 3G difference and not a minor one.

A few thoughts:

1. It's only "required" to be that meticulous if you are trying to conduct a valid DBT that conforms to the protocol accepted by the scientific community, for lack of a better phrase. This does not answer the question whether a DBT is adequate to determine whether there are, in fact, audible differences between two components, for example. I don't want to start a debate on this issue now, but DBT's have certain and procedures, as well as limitations, that may (or may not) impede the ability of participants in a audio trial to hear differences that they might be able to hear under other conditions. And, in any event, it's not required as a matter of law that one conduct a DBT to determine that something sounds better to one's own ears -- unless you wish to impose that condition on yourself.

2. Whether a difference is "small" or not is in the eyes, or ears I guess, of the listener. If one component has some slight sibilance to it, and another does not, but everything else sounds the same, is that a small difference or a big difference? It can be a real big difference if you hate sibilance.

3. Whether a given magnitude of difference justifies spending $X more is obviously an individual judgment. For example, one person might be fine with a set of $100 IEM's with their MP3 player and think it is ridiculous to spend double to get a slight improvement in sound quality. The next person might feel it is worth it to them to spend $700 for top of the line universal IEM's and an outboard amp. Everyone has their own preferences and system of valuation.

4. I would think it would be difficult to judge whether a particular improvement or "difference" is worth it to you until you hear it. I suppose if one lives on the street, once ought to make sure they have food and a place to sleep first before they undertake any discretionary spending on audio equipment. But since most of us on this forum have some disposable income, it's really a matter of what are the best choices for our discretionary spending. And it's hard to make a definitive judgment that some improvement in your audio system (portable or otherwise) is "worth it" -- relative to what else you could spend the money on -- until you hear it. Which is why it's great to be able to buy most products in this hobby on 30-day return.

 

[iriverdude]

Another question is "Do all poweramps, of identical tech specs- power, distortion, VA - sound the same?"

I had two pre-power Hi-Fi systems of different makes, I could tell the difference between stereo analogue pre-amps straight away, but too close for poweramps. But they were close in specs, 60W & 75W, both similar in quality but the 60W generally thought better as matching system.

 

[milkweg]

My Audiolab 8000A and my Onkyo SR505 sound the same to my ears. I did my own AB testing on them one day, not scientific but good enough for me. What surprised me though was that the noise floor of the Onkyo is dead silent and the Audiolab 8000A isn't at full volume setting and that amp was given high marks for quiet noise floor.

Now I have people telling me I should get Marantz 6003 for better sound quality with my Totem speakers but I am sceptical.

 

[iriverdude]

Audiolab 8000A was never very good, mate has owned all Audiolab & Tag gear including grey and black A's- he especially dsliked the grey A's. Try and get pre IAG 8000S or 8000Q. I went from S to Q.

 

[JaZZ]

Quote:

Do all amps sound the same?

Definitely not.

As to power amps, I have two of them and another one in good memory.
The Metaxas Solitaire has clearly the best sound -- subjectively and «technically»: Incredible resolution and airiness, pronouncedly large bandwidth, full, clear, accurate and natural bass, sparkling, smooth and natural treble. Liquid-smooth and airy at the same time. Maybe bordering on euphony in this regard, but not apparently so. The BEL 1001 is quite a bit an antipode to the Solitaire: extremely dry and precise, slightly lacking low-bass full- and blackness, which makes the upper bass catching overly attention, treble bordering on harshness, neverless extended, overall a very dynamic, lively and hot presentation without any affinity to euphony. The Rotel RB 980 was sort of a middle ground: Great, round low bass with good control, smooth, but well-textured midrange, treble with similar characteristic, but lacking the sparkle, smoothness and extension of the Metaxas and the bite of the BEL.

When it comes to headphone amps, the sonic differences are even easier to detect -- not least because swapping them is relatively easy. Just to mention the last three Meier-Audio generations (I've been a beta tester for Jan a while ago): Aria / Opera / Symphony. This order stands for an increase of transparency and resolution. Logically Opera and Symphony are closer to each other (they share the «active ground» concept), nevertheless the improvement is quite noticeable, especially with detail and soundstage.

Now, did I have corresponding expectation beforehand? Yes, certainly. Were my comparisons blind in any way? No. Does this make my experience untrustworthy? It's up to each reader to decide.

If you have decent experience with tube amps, you may have noticed quite distinct differences among them, by tuberolling and of course in comparison to solid-state amps. People who are insensitive to such «glaring» differences disqualify themselves when it comes to high-quality audio in my book.
.