Quote:
Originally Posted by
AJinFLA 
Quite possibly true. Can you supply scientific data to show the correlation between "cheap" parts and audio sound quality? You are implying effect on soundwaves, correct?
Degrade the signal
audibly, or in some other way? Can you supply scientific data to support this assertion? Listening test statistics would suffice. Thanks.
Do you have data to support these claims? How did you do a spatial comparison between your modified and unmodified KEFs? What method was used?
What about ground loop potential from the extra amplifier(s)? No possibility?
How did you separate the modified in-room response being responsible for the changes you heard, vs the mere fact that you went active? I noticed you couldn't supply any data to support your claim of maximum SPL from your KEFs when asked earlier. Can you provide
any data to support any of your claims here? Thanks.
cheers,
AJ
I'll hit this in order:
A. One of the big differences in parts "quality" is the tolerance %. If say one crossover is using 10% tolerance parts vs. another that might use a lower 5% tolerance rating as to how the signal integrity is affected. On simple 1st order corssovers its less critical, but on steeper slope/more complex crossovers with more parts, you now becomes more critical how good the parts are as you add up the compnents used. So say you have a driver that is 3db more efficient than a 2nd one that it will be matched up with. You put a resistor in line to decrease its amplitude of the first, but if the tolerance of the resister used is poor and not spot on to its impedance rating it could mean the output of the 1st driver is a bit too muted or still has too much output compared to the 2nd driver.
On complex crossovers you have alot of parts, all of which have % tolerance ratings........the more parts inline, the greater the chance for signal alteration/smearing from driver to driver as an example.
B. Again, if say between you have 4 drivers(2 each per speaker)on your main monitors with a +/- 3db rating from X to Y frequency response, thats a big window for error at countless points in the frequency spektrum in amplitude in which you can have a 6db swing more at any given frequency. If you have one tweeter thats +3db at 10khz on the left side, but its down 2db on the right side thats a 5db swing. Now say also at 5khz that same tweeter that is hot again at 2db vs. -2db on the other side's tweeter, it can all add up to image being shifted and smeared. When your able to bypass the passive xovers and actively level match each individual driver to each other in room, and then EQ out each individuals driver at multiple frequency points to match them up tighter, you end up with a speaker with a more focues sound and image. Im sure you can understand the logic in this.
C. I have quite a few pair of Kef's, only one pair was modified though(the one I used for listening to music) for 2 channel music. BUt the direct difference in response using pink noise and signal sweeps was significant. I was able to match the left and right speakers/drivers at all frequencies to within 1-3db(+/- .5-1.5db)through the spektrum vs originally a +/- 8-9db in room difference(And the passive xovers alone accounted for a 2db difference on the midrange driver matching L/R and a full db on the tweeters as well). Via use of course driver level matching and EQing. The biggest noticeable difference is the pinpoint placement of sounds, vocals and instruments. Not a small difference, but a starting one.
D. Ground loop issues? Not for my setup.
E. Measured. And repeatable with good measuring equipment(Audiocontrol SA3055), Dayton Omnimic V2 and multiple set up/testing discs Ive had from my IASCA days, along with a multitude of Stereophile's own testing discs and various other pieces of software.