While I haven't done a spectrum analysis, I don't need to; it is simple 'white noise', which has an increasing amplitude as frequency increases, therefore it is heard as coming more from the tweeter, horn or not. Having an efficient speaker is like putting a magnifying glass on the noise coming out of your system. By the way, my woofers are horns too.

[This message has been edited by soundhound (edited November 25, 2002).]

Sorry - but that still doesn't make sense. ALL boats rise with a high tide.

Since posting I made a few calls to recording people - and they could not think how a high efficiency speaker with a FLAT curve (hopefully more a 'line') - and thus a HIGH FIDELITY speaker (fidelity to the original sound - garbage in, garbage out) - would be tilted more towards reproducing 'hiss' (let's call it white noise since we don't really know the spectrum except what's been said here seems to at least indicate a white-sound 'mix') than any other speaker - high OR low-efficiency.

If Klipsch has crazy peaks in the upper (over 5 kHz) area then it is just an anomaly - and however much one might like it - (some people like ketchup on a good steak) - it is NOT a high fidelity speaker! (Mind you I have NOT accused Klipsch of such - I am just saying that would be the logical consequence if one is saying Klipsch is 'hissing' more than other speakers).

[This message has been edited by John Padova (edited November 25, 2002).]

You aren't considering what I said above about 'white noise'. It has an INCREASING amplitude as frequency rises, therefore it has more energy going to the tweeters than to the woofers. If it were 'pink noise' which has equal energy _per octave_ then it would be heard as coming more from _all_ the drivers. Of course all speakers, horn or not reproduce equally across the spectrum (or should), regardless of efficiency, but since the nature of the noise is tilted toward the high end, that is what is being heard.

BTW, my horns are not Klipsch.


[This message has been edited by soundhound (edited November 25, 2002).]

In my understanding on systems like this the older 950 displayed noise that was not gain dependent, thus although the system had lots of excess gain the noise floor was elevated. So in theory all boats rode up, but in fact, unless one can attenuate their ears (and/or have infinite amp power) the practical effect is an elevated noise floor.

The new 950 shows noise that tracks gain if I'm understanding correctly, which is more useful.

Charlie:

You are right about the new 950. Lower the volume control, and the noise lowers also. The 'old' 950 was 'worst case' all the time, regardless of the position of the volume control.

Sorry - dense between the ears maybe here

If a speaker produces a straight line OUT - which I feel a high fidelity speaker should - what difference should the TYPE of speaker do to that signal coming OUT? The speaker should be basically thought of as a black box and what is before the OUT of the box should be irrelevant once we have chosen it to be our reproducer.

Therefore any efficiency issues should have NO effect on the spectral distribution as simple volume adjustment should make OUTPUT 1 = OUTPUT 2.

There is the possibility that one MIGHT listen to horns at HIGHER volumes than non-horns because lower frequencies might have more 'muscle' subjectively - thus 'filling' the ear from the less efficient (non-horn) speaker at lower output levels. But you say you have high-efficiency horns at lower frequencies too.

But - really - without trying to be confrontational in an impolite way - I am still stuck on the 'all boats/high tide' analogy.

[This message has been edited by John Padova (edited November 25, 2002).]

You are absolutely right about what you say. Efficiency has no effect on the frequency balance of a speaker. Straight line in, straight line out. BUT the signal (noise) 'in' in this instance is tilted toward the high end of the spectrum, and the speaker is just passing along what it is getting; noise that is accentuated in the high end. Horn or not, you'll hear more energy from the tweeter than the woofer with this type of noise.

If you've used a spectrum analyzer to check the frequency response of a speaker, you've undoutedly used 'pink noise' as the signal source. This type of noise should display a flat line across the frequency spectrum on the analyzer. It sounds just like it looks, equal across the band. "White noise" is never used as a signal source, and if you looked at it on a spectrum analyzer, it would be a rising line from low to high frequencies. It sounds just like it looks. White noise is the type of noise that is generated in electronic equipment. Pink noise is artificially produced by imposing a 3db per octave filter on a white noise source.

[This message has been edited by soundhound (edited November 25, 2002).]

Does that thus mean you are of the same opinion that the 'type' of speaker reproducer is irrelevant - if each type has the same spectral effect (hopefully none- but that is also irrelevant) on the input signal?

This should also mean that the output SPL's are also equal because the ear is more sensitive to lower and upper frequencies at higher volumes (our friends Flethcer and Munson - who have, btw, opened a chain of steak houses - alas - no ketchup).

Technically, that should be the case. I practical terms, it is not. Horns just sound different than direct radiators. The dispersion patterns are very different between them for starters. Horns also can in theory produce lower distrotion by virture of the fact that the diaphrams don't have to move as much to reproduce the same sound level. Horns can also handle sound pressure levels that would shread a direct radiator speaker. Horns have their own set of problems. They can sound colored of not implemented just right.

In a nutshell, pink noise is distributed such that the same amount of energy is in the space from 100-200 Hz as in the space 10000-20000 Hz, whereas white noise has constant density, such that the same energy is present from 100-200 as from 10000-10100.

If I understand correctly:

The issue with high gain amps/speakers is that if (as was the case with the 950) the noise floor is a fixed level and the gain from the pre-out to ear is high, this noise floor will rise to an annoying level even though the gain is turned down. Any adjustment in that chain (amp gain for instance) can be used to bandaid it, but the new revision sounds like it's correctly implemented.