Originally Posted by Brian Owens
My mains are more than loud enough. They are extremely loud, maybe too loud for that size room. And the scan speak drivers do not seem to have any problem with distortion, even with the high power levels.
I guess you just have to experience it to understand. . .
Hats off to ya again. WRT your mains, if frequently posted about the subtle effects of compression in cone and dome based systems...it's very easy to enter into significant compression artifacts without knowing...because of the nature of compression distortions are not offensive like other types of distortions.
Scan Speak makes some of the finest drivers in the world....doesn't matter. Please, do not take offense, this thread and your build are just killer, and everyone around is off the chart jealous. The peak demands of H/T material, and typical cone/dome driver speaker systems, regardless of pedigree, just is a comprimise.
Here's my take on compression;
Mixing live, in every situation imaginable, has taught me how to listen very closely as a system of any size reaches it's first weak link, which I typically find is compression.
It's easy and one needs to train themselves to notice the onset of thermal compression. Heat, a by product of current flow, affects the individual drivers and generally sets in before the onset of clipping. Clipping, even moderate clipping, really accelerates the effects of thermal compression, and the combination can lead to driver failure. Loudspeaker drivers are very inefficient to begin with. The amount of power (approx 98%) that is wasted by heat in the voice coil/motor assembly is significant. The manner in which this rise in temperature is handled and dissipated is what is critical. Vents around the motor structure, the former material and high tech adhesives, all play a role in addressing the thermal characteristics.
As the voice-coil heats up, its resistance increases. The driver's sensitivity decreases with this heat, and as you drive them harder, they become less responsive. In my experience, thermal compression can be described as a subtle thickening, or congestion of the output. This causes the presentation to lose the snap, and the nice leading edges to the transient detail. Also, the LF roll off is determined by many things, one of which is voice coil resistance. So the LF performance characteristics change in addition to the blunting of transient detail. Thermal compression can easily change the tonal balance of a crossover based, 2 way or 3 way etc. system. The driver in the speaker with the most significant compression changes, and lowers in volume compared to the other drivers. So not only does it lose the transient detail, the tonal balance shifts.
The linear nature at which sounds increase in level in real life, is what one strives for with regard to system dynamics. While the amplifier and speaker are working within their linear envelope, the amp isn't clipping and the driver isn't exhibiting significant compression. Each rise in level demanded by the source signal, or by the volume knob, is tracked relatively accurately by a subsequent rise in amp power. In turn, the drivers operating within their linear thermal envelope, react proportionally and retain a nice dynamic quality. Now when either the amp clips, or the speaker can't thermally dissipate the heat quickly enough, the dynamic non-linearities become greater and greater. If the amp has enough power, then one of the speaker sections begins to give up first, say the LF. So the bass guitar loses articulation and becomes muddled. The kick drum softens and loses that nice snappy leading edge to the boom. The crossover networked that was painstakingly voiced, after substantial computer modeling by the designer, changes characteristics. Where transients should jump out at you as they do at a live event, they lay flat across the mix, without life and any impact.
Thermal compression is obviously dependent on a buildup of heat. A rarely discussed different form of compression is flux modulation based dynamic compression. This also causes causes a dulling of transients and distortion is created. As I understand it, it's an instantaneous magnetic overload, whereby the magnetic circuit can't sustain the maximum flux across the gap as the signal level changes. This magnetic compression causes similar effects as thermal compression, but their onset has no correlation to voice coil heat, just an inherent design inability to handle the largest peak. Little information with regard to magnetic compression is available in speakers, however, oftentimes speakers are afflicted with at least 3 to 6 db of thermal compression. That means easily half the impact, and as high as 75% (6db) of the power is diminished due to thermal compression. This is huge!
Removing all the compression artifacts would be great. Like electro-stats, they don't have these problems. And if you've ever heard electro-static speakers, you most likely heard the nice quality of not having the same negative magnetic effects that dynamic speakers have to deal with. But stats have their own problems. What they do, they do extremely well. Their just not well suited for high output, high performance H/T. To alleviate the most negative artifacts of dynamic compression, you need relatively high sensitivity, high power and output capability, and perhaps some measure of controlled directivity in a dynamic driver based system.
Speaker sensitivity, is very important in high performance H/T.
Please, no offense intended, your rig is killer! You know everyone's drooling over those pics, showing them to their spouses etc.. Just for information, many don't consider compression due to the subtle onset.