How are speakers improved?


Lets look at a hypothetical example of a speaker designer. Lets call him Nigel. Now Nigel has made a bookshelf speaker. It uses state of the art drivers and a braced MDF box. The crossover is 4th order linkwitz riley @3khz. All very typical so far. The speaker is almost ruler flat within its operating range. Off axis is even and quite flat. Distortion is within typical limits. 

He listens to it and sends it off to some reviewers. The response he gets is lukewarm. Its neither good or bad. 

The questions are:

1.) What does Nigel need to do to IMPROVE the speaker beyond its existing performance?
2.) What parameters need to be improved in order to elevate the performance to an even higher standard? 

WHAT IS THE SECRET TO PERFECT SOUND THAT HAS ELUDED NIGEL???
kenjit
Just want to say how frustrating it is to see a title and it seems good and so you click only to realize aw dang its him again. What a waste. That's all. Bye.
Nigel needs to install furniture shakers so that he thinks he did something different than everyone else
1. Piano black and 2. quadruple the price.
That would certainly trick audiophiles into thinking there had been an improvement. But it wouldnt be real. 


I don't know who this Nigel character is, but I can say the furniture shakers are a huge hit with the ladies.
kenjit the answer is obvious Nigel needs you to come and tune his speakers for him.
Problem is Nigel can’t hear for beans and has been duped.  Shoot, Nigels speakers have been out of phase with one another for the past 10 years and he didn’t even know it.


Nigel should tie a bose radio around his neck and jump off a bridge.


Back to the music, ta ta for now!
Nigel is employed by a corporation and he is looking to develop a bookshelf speaker to compete with Brand X, only a better sound at the same price. After looking at some driver specs, he sees something interesting: if he fiddles a little here and there; uses this particular driver and reduces the crossover to 2K he can produce that speaker and get the better sound at the price point.

Nigel sends the design documents upstairs. A few days later he gets a call from Engineering saying great job! But... make one small change, if you don’t mind. Increase the crossover frequency above 2.8K.

Nigel is disappointed -- the crux of the design is that lower crossover frequency. Nigel asks why and Engineering informs him that they contacted the driver manufacturer and the verdict was the longer driver excursion at the lower frequency will increase MTBF by over 45%. The accounting department determined that the extra warranty cost involving labor, stock of replacement drivers etc, will lower the ARR to less than 3% over 10 years, so the project isn’t feasible. The price increase to be doable will price it out of the market.

Morale of the story: if it’s not your money then you don’t call the shots.


Brother K, you asked a question for Nigel, so I'm gonna answer Nigel. 

Pay the reviewers to LIE.  Now that you know the secret of Nigel's disappointment and a solution to his ultimate success, I hope you move in that direction.. 

In other words... Save up some cash for Nigel, probably the best idea.

Retired Teamsters L 315,

cheat um if you can...rob them if you can't.. "Billy the Kid"

Regards
The very notion that a designer starting with state of the art drivers can achieve perfection has an implicit "begging the question" problem--it assumes that there is such a thing as a state of the art driver and that driver has the potential for delivering perfect sound.  Even taking "perfection" out as a goal, there is no such thing as drivers that represent state of the art in all aspects of performance; there are only drivers that are good at certain aspects of performance and remain workable in those aspects that they are not that good.  Sometimes, the very thing that makes a driver a contender for state of the art in one aspect of performance (such as the dynamics as transient response of the Manger bending wave drivers) makes them particularly difficult to integrate with other drivers in a coherent fashion.  The craft of designing and building speakers is not simply putting together the "best" this or the best that, it is almost the opposite--it is working around and ameliorating the problem area of each component.  Even when that is done well, a particular speaker will have inherent limitations--it will sound its best only in a particular room location, it will sound good only to someone who shares the designer's taste and priorities, etc.  

There is no such thing as a consensus choice of a "best" crossover design--the best design is a matter of taste and priority, the characteristics of the driver, the characteristics of the speaker cabinet and a whole host of other issues.  Also, many crossovers are not merely dividing the frequency spectrum between drivers, they are correcting for frequency response issues, driver phasing, "baffle-step" issues, floor bounce issues, etc.

The best cabinet design is, again, the best for the particular drivers, the best for what the designer is trying to achieve sonically, and that does not necessarily mean something braced so well that it is sonically dead.  If that were the case, designs like those of Harbeth and Audio Note would never sell.

What Nigel needs to learn is that design in not simply a matter of taking the best from column A, B, and C to slap together a speaker.  He must go the much harder route of determining what sound he hopes to achieve (with the understanding that MANY compromises and trade-offs will have to be made); experimenting with the almost infinite combinations of drivers, cabinet designs, and other design factors, and finding a way to manufacture the final design achieved after all that hard work.  
The problem with well thought out, educated responses like larryi and gs556 is that all they are doing is educating the troll so he can come back next week and start the same thread only now armed with more knowledge (notice I didn’t say experience)...
We're Only Making Plans For Nigel.
Nigel can be found most days at the Nigel User Group  (NUG)
YUP... I personally see the change in response, thing are a changing..

I feel a sonnet coming on..:-) Why not enjoy it.. LOL, HO HO HO...

Confucius say "man who eat jelly beans Farts in technicolor". Old but true..

Romp....

Happy Happy, Romp, Romp, Romp, Romp, Romp.. rainbow... right.. :-)

Regards
Yeah, same here...haven't listened to Drums and Wires in years. Nigel triggered my memory. Gonna cue it up tonight.