What is "slope"???


I dont think I totally understand the properties of slope?

I'm setting up a Theta Casablanca and when it comes to crossover I get it,....but as far as setting the slope, Im at a loss as to what it addresses.

Please help me understand what I should be listening for.

Thanks.
mhubbard
"Ezmeralda, Jeffrey -- thanks for the the clarification on the slope values. My memory doesn't seem to work quite as well as it once did ;-) Next time I'll confirm before I post."

You're welcome. Happens to us all. I had a poor memory when I was young, and I think as I got older, it got worse...but I can't remember. Now at 60, I have too many 'senior moments'.
Ezmeralda, Jeffrey -- thanks for the the clarification on the slope values. My memory doesn't seem to work quite as well as it once did ;-) Next time I'll confirm before I post.
Does the Casablanca give you the choice of sending the full signal to the subwoofer channels (i.e. SLOPE OFF)? If so, then why not set the subs to receive the full signal and let the sub's filter let in what it needs.

But if you want to bypass the Thiels completely, start with the 80 hz frequency at 24 db. This steep roll off may result in a "hole" at 80 hz. Keep going down the slope decade until the sound is coherent, that is, the sound levels at that frequency are similar between the sub and the speakers. Ideally, they should be within 3db at 60,80 and 100 hz but not more than 6 db. A sound pressure level meter will greatly help. The bass should sound fast and quickly disappear - not resonating or reverberating.

If you go to 6 db, then the speaker woofer may overrun the sub and you'll get reinforcement (louder, bass boom, fat bass, etc.) at that frequency. Sorry, but this is probably a trial and error process. Also note that the Thiels have a 6 db slope - but that is with its own drivers and will probably not work so great with a sub at that slope setting.

BTW - make sure your speakers are properly set up before trying out the different slopes or you may not notice a difference.
So you're talking about the high-pass filter to the Thiels?

If so, my 1st response is 'why bother'? Using any of these filters introduces phase errors into the bass and midrange frequencies. Also, it sounds as if you're using one subwoof, so filtering the bass from the Thiels makes the low- and midbass mono instead of stereo.

My 2nd response is the less 'slope' (and phase errors) the better. If you feel you MUST use a hi-pass filter, use as gentle a one as you can select.

But try running them fullrange and filling just the bottom octave with the subwoof.
My speakers are full range Thiel 2.3's. They are capable of full range bass, but I cross them over at 80Hz to my sub. I find it allows the Thiels to do what they do best.

The slope options available are 6db, 12db, 18db and 24db

Hope this helps.
M, is this a high- or low-pass filter? Do you have large, full-range speakers or ones with relatively limited bass output? Do the main speakers have plenty of power-handling capability full range or can you overdrive them at YOUR normal, high listening levels?
OK, so the second part of my question was..

What should I be listening for to establish the proper slope for my processor?

Please be as straight forward as possible.

Thanks for the help!
I think 1st is 6 db (decibal) an octave roll-off, 2nd is 12, 3rd is 18, and 4th is 24db an octave roll-off. That's the slope at which the signal is rolling-off With 90 degrees phase shift on each one going up also.
"Expanding just a bit on Marakanetz's response, the slope is the decrease in signal strength per octave. So a first order cross-over has a slope of -3db and it goes up from there -- second is -6, third is -12, fourth is -24. The decrease begins at a certain frequency and keeps on going from there depending on the specifics of the cross-over."

Well...not quite. First, 'slope' references the steepness of the plotted falloff of signal strength v. frequency. When someone says 'the slope of the filter', they mean the steepness of the filter when visualized (= plotted on paper). But that term is rather casual and inexact.

A filter point is defined as that frequency which is 3 decibels (dB--the 'B' is capitalized in the abbreviation because it was someone's name) above or below the zero or reference level. A 1st-order filter has a 'slope' or steepness of SIX dB per octave. That means that with a perfectly behaving low-pass filter, the signal strength one octave (= double the frequency) further up the frequency scale is now 9dB below reference level. The signal one octave higher yet (= another doubling of frequency) is another 6dB down, or a total of 15dB, and so on. For example,with a 1st-order low-pass filter of, say, 100Hz, the signal strength will be down 3dB at 100Hz, 9dB at 200Hz, 15dB at 400Hz, 21dB at 800Hz, etc. A 2nd-order filter has a slope of 12dB per octave; a 3rd-order, 18dB per octave, and so on.

In a 2nd-order filter at 100Hz, 100Hz is still down 3dB, but 200Hz is down 15dB, 400Hz is down 27dB, etc. With a 3rd-order, 200Hz is down 21dB, 400Hz is down 39dB, etc.

Higher-order (means 'higher than 1st-order') filters get rid of 'unwanted' signal to a greater degree than 1st-order filters; they're often used to increase power-handling capacity of a driver and hence a system. For instance, if one had a woofer that sounded really poor in the lower midrange because it was slow to respond to faster signals, one could use a 3rd- or 4th-order filter to keep more of the midrange out of that woofer. However, higher-order filters have increasing amounts of phase error, so there's a price to pay.
Expanding just a bit on Marakanetz's response, the slope is the decrease in signal strength per octave. So a first order cross-over has a slope of -3db and it goes up from there -- second is -6, third is -12, fourth is -24. The decrease begins at a certain frequency and keeps on going from there depending on the specifics of the cross-over.
If you see a plot of freequency v.s. signal when it comes to crossover, at some point which crossover is designed or multiple of such it should have a slope(s).
Since the freequency can't be cut of perfectly vertically there slope defined exponentially.