[Construct]

The logical mind needs transparency and linearity, but the soul requires character, color, and personality.

Good way to break it down.  That is why I can enjoy *any*  speaker presentation depending on mood/state of mind.  Sometimes I prefer realistic spl and impact, other times I want that almost ethereal transparency so it put's me into the music, vs in front of it.

[Aom-Uom]

cujobob,
class AB switching provides efficiency by reducing power loss. Crossover distortion is very unpleasant, different to the "coloration" in loudspeakers, which is a blend of high level harmonic distortion, intermodulation distortion, frequency distortion, and phase distortion.

[Construct]

Of course tests have shown inability to tell the difference between amps...let alone someone stopping to gasp in the middle of an a/b/x: "Gahh!!!  That's class A/b crossover distortion!!!!  The sound is ruined!!!  Put the pure class a amp back on!!!"

[Aom-Uom]

Construct,
now you are moving in the right direction, will need a big fat heat sink...

[Construct]

Construct,
now you are moving in the right direction, will need a big fat heat sink...

I seem to remember hearing a pure class a amp or two in my day.  A lot of them tend to be romanticized, and sacrifice bass grip, dynamics and slam. (single ended types etc) Oh well, they are good for those intimate cold winter days sipping brandy in an easy chair listening to baroque. At least you have no need to start the fireplace going for heat. 

[drew54]

interesting discussion, in part cause no one defined "coloration."  I will define it as follows:  Any departure in tonality from the source.  Dynamic compression is therefore not coloration. 

But...
Lack of bass or hi treble (single drivers), shouty midrange (single drivers and some horns), midbass suckouts (my opinion of some highly regarded spkrs including the big Quads and Vandy 5s)
are all colorations I choose not to live with.

Hi sensitivity does not necessarily mean high coloration.  But corrective actions to remove colorations (complex x overs, enclosures, multiple drivers etc) all come at a cost of sensitivity.

Tonal neutrality does not guarantee a great speaker, but sure is a better starting point than obvious deviation.

Finally, two of my favorite systems owned by friends are based around 1) well driven Klipschorns (104db sens) and 2) Von Schweikert VR-6s (94db sens).  It can be done, its just not easy. 

[jriggy]

interesting discussion, in part cause no one defined "coloration."  I will define it as follows:  Any departure in tonality from the source.  Dynamic compression is therefore not coloration. 



Hi sensitivity does not necessarily mean high coloration.  But corrective actions to remove colorations (complex x overs, enclosures, multiple drivers etc) all come at a cost of sensitivity.

Tonal neutrality does not guarantee a great speaker, but sure is a better starting point than obvious deviation.

Thank you very much drew54!

 ~J

[Aom-Uom]

drew54,
the definition, distortion is any departure (except magnitude) of the output signal waveform from the input signal waveform, should give you a solid ground. I prefer to use "distortion" instead the vague term "coloration". The distortion types occurring in loudspeakers are enumerated in post #25. You will unlikely get closer to the truth.
Yes, to achieve high sensitivity you must employ methods that increase distortion. There is no way to get around this relationship.

[Aom-Uom]

Quiet Earth,

I would bet real money that we all prefer a little bit of compression because that's the reality that we live in. Someone recently mentioned (in another thread) that the ear compresses sound anyway.

The ear acts very much like a Klipsch horn in a way, so if you could build a Klipsch horn with the ear´s characteristics, no "coloration" would be heard. A compression ratio of 4:1 will be fine.

[Aom-Uom]

Construct,

A lot of them tend to be romanticized, and sacrifice bass grip, dynamics and slam. (single ended types etc)

Standard discrepancies may arise if not everything is taken into consideration.

[cujobob]

I would bet real money that we all prefer a little bit of compression because that's the reality that we live in. Someone recently mentioned (in another thread) that the ear compresses sound anyway. I have no clue if that's true but I know that a little compression goes a long way when you make a recording.

The logical mind needs transparency and linearity, but the soul requires character, color, and personality.

I would disagree with this.  I now use a pair of Gedlee Abbeys and one of the first things I noticed by comparison to my GR Research OB7s (also very fine speakers) was that the Abbeys had a more lifelike, uncompressed sound.  There are waveguides/horns, and waveguides/horns done well.

Aom-Uom, I realize the advantages to Class AB amps, but the disadvantages are certainly worth noting.  While crossover distortion is indeed different from loudspeaker 'distortion'...it's a big issue that is mostly ignored.  Inefficient speakers require more power and with that comes its own set of flaws.

Constant directivity is important for good sound, too, and that's hard to get without a waveguide...found in a speaker with high sensitivity.

[Jon L]

IMO High Eff doesn't *have* to be more colored than others.  However, high efficiency drivers tend to trade efficiency for bandwidth (and or linearity at bandwidth ends).  Therefore, if one constructs a high eff speaker yet try to extend bandwidth of those hi eff drivers (not using enough drivers to cover ranges they are comfortable in), bad things can happen.

Some try to get around the issue by mixing totally different technologies, i.e. ribbon tweeter with large woofers, compression tweeter with dynamic woofers and/or ribbon, etc.  More power to you if the designer's skills prevail in these situations 

[cujobob]

Spend enough money on drivers and you can get around most of the problems that plagued them in years past.

[TRADERXFAN]

-and don't forget the minimal diffraction on those Gedlee's... that is a significant factor.

[Aom-Uom]

cujobob,
you have apparently misunderstood me (which is by no means your fault), I have stated on several occasions, that switching distortion is the worst distortion in audio and linearity is inversely proportional to efficiency.

Inefficient speakers require more power and with that comes its own set of flaws.

It´s very true, as distortions rise with increasing output power, but generally, loudspeaker quality is more decisive for the perceived sound.

[Quiet Earth]

Since this thread is spinning between amplifier theory and speaker issues, I'd like to spin it some more  . . . .

I would rather have a fabulous sounding amplifier driving the most average speaker instead of a real boring sounding, power house of an amplifier driving the finest speaker ever made. Color me crazy!

[Duke]

[D]istortion is any departure (except magnitude) of the output signal waveform from the input signal waveform...

Whether it fits the definition of "distortion" or not, thermal compression is an audible and undesirable characteristic exhibited by many loudspeaker systems.   

[T]o achieve high sensitivity you must employ methods that increase distortion. There is no way to get around this relationship.

To the best of my knowledge, increasing magnet strength raises efficiency without increasing distortion; in fact, an excess of magnet strength ("saturated pole piece") reduces flux modulation, and thus reduces some kinds of distortion.  Another techinque for raising efficiency is increasing the suspension system compliance, which does not in and of itself increase distortion. 

Duke

[BobRex]

Since this thread is spinning between amplifier theory and speaker issues, I'd like to spin it some more  . . . .

I would rather have a fabulous sounding amplifier driving the most average speaker instead of a real boring sounding, power house of an amplifier driving the finest speaker ever made. Color me crazy!

Actually, it's not that crazy.  Way back when I sold audio we used to take some of our cheap speakers back into the "high end" room and give them a shot.  Imagine a pair of Fried Qs tied to a cj Premier 1 amp.  The results were always impressive.

[SET Man]

Hey!
 
    Interesting topic here. At least the OP used "Most" in the headline.

   As for hi-eff speaker. Well, there are tread off with this.... eff vs freq extension. Most flat respond speaker that I've seen tend to be low-eff. But than again you can have flat respond but there will be some coloration to the sound.

    Anyway, I'm surprised that Magnepan is mentioned here as low-coloration speaker. Guess what? I used run Maggie 1.5 in my system until built myself TQWT based single driver with 6.5" Fostex driver. And all I can say is that Maggies are now store behind them acting like a tuning sound panels.

    There is one thing that hi-eff have that low(lower)-eff speaker don't have and that is wider dynamic swing respond. And this really make thing comes alive. I'm not talking just dynamic respond but all this also effect the finer detail of sound as well even at lower listening level. Hard to explain.

     The low-eff Maggie sound good too but need to be play at higher level with more power even that they still don't have that finer dynamic swing to the sound. And of course there are dipole and this make them special to some people.

     Anyway, I don't know where to go with this. It is true that all speaker have colorations, some are more than others. But the trick is to choose one that doesn't bother you. It is all depend on your own preference I guess... what you value most in sound.  So, just choose the color that fit you best

Take care,
Buddy

[Aom-Uom]

Duke,
thermal compression largely affects linearity. Very unfortunately, under constant voltage drive, the voice coil resistance is forced to act as a (poor) current source for the (inappropriate) voltage-to-current conversion. (Explanation may be necessary here).
Magnetic flux clearly raises efficiency, because the force generated is the product of the magnetic field strength multiplied by the current through the coil.
Flux modulation is a nonlinear interaction between the static magnetic field and the AC field generated by the voice coil current. Flux modulation increases both with increasing flux density and increasing AC current, causing high frequency harmonic and intermodulation distortion.
Linearity in loudspeakers depends on the delicate balance of a large number of (ubiquitous conflicting) parameters.

Please keep trying to disprove my statements.