Originally posted by wolfgang
To be honest I would be more interested to know if any professional manufactures have conducted listening tests based on some form of DBT since they are more likely to have the means and capability to design one of any scientific value.
I found that a very interesting and thoughtful post. Thanks Wolfie. As a punishment, here's a longish posting of my own.
I'll try to answer your question above, as it applies to my company.
As a manufacturer of speakers, I confess that I haven't yet conducted any DB tests. I have run a number of single blind tests for various A/B comparisons.
Sometimes when I make a change, the results are stunningly obvious to me (I design for myself). If option 'B' seems clearly better than 'A', I'll live with 'B'. After some days, I may continue with 'B' or, if further listening has shown up details that I hadn't originally picked up. I may compare it again with 'A' to see if my conclusions remain the same.
If I make a change and the differences are not quickly apparent, I may discard the change or, if I hear that there's something there but it's subtle, I'll make a blind test. Often that would be carried out with my wife and usually we would swap roles as listener and tester.
The listener is blind to the changes; the tester isn't. It's possible that the tester can convey something unintentionally by the listener's subconscious picking up of signals below the threshold of conscious awareness. Perhaps telepathy plays a role. Either way, this is about as controversial as directional cables. If the difference I'm testing is as subtle as the subconscious pick-up between tester and listener, I probably won't bother with it anyway.
Most of my real learning takes place over several days of listening.
There are exceptions to the above.
Setting up the speakers in a 'new' room (eg. at someone's home) involves positioning the full-range speakers, positioning the subs and experimenting with the floor/speaker interfaces.
Setting up the speakers for an exhibition involves all of the above plus playing with equipment, supports and cables.
In someone's home I will do the obvious stuff and arrive at a consensus with the other parties for the more subtle details. There may also be the occasional, impromptu blind test (as you do). Generally, at an initial listening, one doesn't achieve the level of detail that can be achieved while living with the system for a while.
For exhibition set-ups, I'm often setting up in the company of other professionals. For example, at the Heathrow show we will have the top man from Michell, the boys from Voodoo Isolation, Graham of Trichord (who will doubtless turn up at some point during the show), Tone (w-m) and me. Gary of Border Patrol will be away this time; otherwise he would join us.
We each tend to do our own thing with our own pieces of equipment and then do the positioning, equipment integration and tuning. Generally we will work by group consensus, although I tend to give my own views more weight - well I'm paying for the room and it's got the name of my company on the door.
Even during an exhibition there are often a number of blind 'mini tests' that take place. Suffice it to say that near the end of the show we would expect to achieve our best sound.
Of course all of this is independent of the objective tests that come from measurement. Measurements can throw up resonances, overhangs and frequency anomalies. These can then be addressed and the speakers then retested. Most equipemnt companies use meaurement and mine is no exception here. Subjectively, I generally find that a measured anomaly should be 'under addressed' rather than 'over addressed'.
The design process itself goes beyond objectivity, measurement and calculation and into the realms of imagination, feeling and intuition. Any hi-fi designer who has not developed these aspects of himself will turn out, at best, pretty, formulaic boxes. He's unlikely to be making music.
Remember the electrical energy does not exist in the 'wire'. 
