Thread: Filter / Crossover differences - Not all crossovers are equal

IIR (infinite impulse response)

• Explanation
Based on analog LRC crossover circuits and then modeled in mathematics using discretization or approximations. Quantization comes into play when using a computer. Since it has feedback controlling it, the response to an impulse goes on indefinitely (rounding errors of the approximations). Instead of using decimal numbers it uses fixed numbers (i.e.-1,2,3 instead of 1.234, 1.9845, 3.1467).

Discretization – using mathematical formulas to create models of an analog circuit

Quantization – rounding decimals to whole numbers for example an analog to digital converter. There exists debate about this between 16 bit and 24 bit audio and the ability to distinguish between quiet/soft passages of music and louder counterparts. The overall loudness of newer music has begun to increase due to portable devices, so the varying levels of sound are not as important as older music was once mastered. This is what is considered the “loudness war”. The dynamics of newer recordings are much less than recordings from years ago.

Phase shift of 90 degrees for every 6db/octave
Chebyshev filter, Butterworth filter, Elliptic filter

• Advantages
Fast and Cheap
Easy to implement due to analog counterparts

• Disadvantages
Poorer bandpass filtering
Stability issues
Phase shift
Rounding errors constantly increase due to feedback or feeding the rounded value back into the system.

• Programs/Software/Hardware using this type
KX project
Foobar DSP
Most car audio crossovers
Frequency Allocator

FIR (finite impulse response)

• Explanation
There is no feedback controlling it so when it responds to an impulse (short burst of sound/energy) it eventually settles to zero.

• Advantages
Are inherently stable. This is due to the fact that all the poles are located at the origin and thus are located within the unit circle.
Require no feedback. This means that any rounding errors are not compounded by summed iterations. The same relative error occurs in each calculation.

Can have linear phase
Can have minimum phase

• Disadvantages
-Requires a great deal of resources to make an FIR filter and they have fixed values.
-Pre-ringin can occur in frequencies at 3000Hz and above.

• Programs/Software/Hardware using this type
Not FIR but Frequency Allocator+Frequency Arbitrator maintains correct linear phase
BruteFIR - Better have some programming skills.
AKAI PRO Quad Comp VSTplugin- more of a compressor plugin but might work as crossovers with multiple instances.

Examples go here

and here

and here too.

Great thread.

Problem with phase linear crossovers is "pre-ringing", most of the time this not audiable, however hit 3000 Hz and above and your "asking" for trouble.

Looking for the perfect active XO for High End audio speakers there are some basic requirements that have to be considered:

1. A perfect XO has good attenuation outside the pass band to protect the tweeters and suppress irregularities of the chassis outside their optimal operating range.
2. A perfect XO has no ringing.
In the critical region (about 3000Hz) ringing is clearly audible.
3. A perfect XO is transient perfect.
Most researchers argue today that transient imperfect XOs are only audible at low frequencies but there they definitely are audible.
4. A perfect XO has minimum delay.
It should be usable for audio and video applications
5. A perfect XO has excellent polar response.

http://freerider.dyndns.org/anlage/M...eo-Systems.zip

Time alighment

Implementing the “The ‘Perfect’ XO for High End Stereo Systems” described in (1) special care has to be taken of the speaker’s time alignment

http://freerider.dyndns.org/anlage/M...-Alignment.zip

Excellent! I will add this info in about the ringing. I am still gathering information so it will take some time to get this organized. I just thought I would get it started.

Also, something is wrong with your links.

Passive versus active filters..

Passive is as the name implies, is a filter reacting directly with its source. That is it is effected by driver impedance and reactance. There is a rather large band of people who swear by this as the ultimate solution. For us it is simply the cheapest(normally supplied by the manufacturer) and easiest to setup.
They provide "reasonable" driver protection from pops etc.

Caveats are
1\It requires amplifier power to run it, (up to 6 dB loss over active)
2\ Picks up every bit of inductance and RF going. see alt whine that I dont think was there before, with amps off even!
as an example.
3\ Individual drivers cannot be time aligned.(OK yes their are passive delay circuits, but in the real world)
4\ lack of flexibility, these are designed for one set of drivers only. Change the driver and the parameters change.
5\ If one driver is shorted or forms an open circuit (wire falls off, coil destructs) it will present a short circuit to the amp ... at the particular LC resonant frequency of the crossover. (Usually 12 or 24 degree slopes)

Active
Require a external power source to operate and are generally before the amplifier. (Inbuilt into some amps are active crossovers)
Advantage is flexibility in driver choice.
Easier on the amp (doesn't have to power the passive crossover) so possibly less distortion depending on the crossover duties. Plus the amp is dedicated to a smaller frequency bandwidth.

caveats
Expense
Knowledge of crossover filters for the drivers.
No protection for drivers for pops etc.

Originally Posted by durwood

M10000-Liliput 7"-CSB Live USB-Keyspan-Seagate 250mb-Liteon DVDRW USB

What OS you runnin on that?

Windows xp pro...not optimized. One of these days I will get around to removing what is not needed using nlite and a re-install. Also using Roadrunner as my frontend

hahaha...its from the old days 1992/1993...just kidding. I don't know what I was thinking. Fixed. thanks for pointing it out.

Frequency Allocator uses IIR filtering only, however when teamed with frequency arbitrator the phase is supposed to remain intact and speaker frequency response is unaffected. These two programs are looking very promising! Now back to the software thread.