Crossover: Difference between revisions - Bose Portable PA Knowledge

Line 1:

~~quote:~~ I'm familiar with the term 'crossover,' but not really with its meaning.

; I'm familiar with the term 'crossover,' but not really with its meaning.

''Audio crossovers are a class of electronic filters designed specifically for use in audio applications, especially hi-fi. A commonly used dynamic loudspeaker driver is incapable of covering the entire audio spectrum all by itself. Thus, crossovers serve the purpose of splitting the audio signal into separate frequency bands which can be handled by individual loudspeaker drivers optimized for those bands. A combination of multiple drivers each catering to a different frequency band constitutes most hi-fi speaker systems. An audio crossover may also be constructed mechanically and is commonly found in full-range speakers.''

Line 27:

[[Bass Frequencies]]

=== More Bass Talk ===

Hilmar-at-Bose explained in [http://bose.infopop.cc/eve/forums/a/tpc/f/3976055944/m/6121039062?r=2641067362#2641067362 More Bass Talk]

==== Crossover ====

If there is no B1 and nothing connected to the Bass Line Out. The L1 sees frequencies from 110Hz up. Feeding it anything lower, doesn’t make sense, since it couldn’t produce any acoustic output and if would rip the drivers to shreds.

In any other case the L1 sees signals only from 180Hz up. There is no other variation in frequency or gain for the L1 no matter what else happens

==== Bass Line Out and B1 behavior ====

This is based on the design goal that “You should always sound the same; no matter how much Bass stuff is attached” I can try to explain my view of why this is a good design goal (of which you may disagree) but let’s look at the actual behavior first.

===== Without Bass Line out =====

1xB1: 40Hz-180Hz, B1 specific EQ, some nominal gain that we call 0dB

2xB1: 40Hz-180Hz, B1 specific EQ, -6dB as compared to nominal

===== With Bass Line Out =====

0xB1: 40-180 Hz, flat, roughly the same gain as 2 B1

1xB1: 40Hz-180Hz, B1 specific EQ, -6dB as compared to nominal

2xB1: 40Hz-180Hz, B1 specific EQ, -12dB as compared to nominal

What this complex behavior does is the following. No matter if you attach 1, 2, or 4 B1s, you will get pretty much the same balance between all combined B1s and the L1s. It’s a little off for 3, 5, 6, 7 & 8 B1s, but still reasonably close.

==== Frequency content of an acoustic guitar ====

Oldghm, you did some really interesting experiments there. However, you have to be really careful when using an RTA. You can feed these things a pure sine wave at 80 Hz and by turning it up make the 63 Hz and even the 40Hz LED light up. They will be lower than the 80 Hz LED, but still come on. That does NOT mean, that the sine wave contains any other frequency than 80 Hz (it certainly doesn’t). It only means that the RTA has a pretty limited frequency resolution. The 63 Hz LED will respond best to 63 Hz signal but it’s in no way “blind” to 80 Hz signal.

Thus being said, the actual frequency content is not easy to determine. All sounds that have a pitch are certainly constraint to 80 Hz and up (in standard tuning) and there isn’t actually too much energy at the fundamental. However, the “non-pitched” sounds like a hard string attack or whacking the top with your hand can very well have lower frequencies. Unfortunately, I don’t have any hard data on that, but we will measure that at some point.

==== Equal loudness curves ====

Here is the bunch

http://hyperphysics.phy-astr.gsu.edu/hbase/sound/eqloud.html

These curves tell us two things:

First, the same physical sound energy produces different perceived loudness depending on frequency. You can turn that around into “The physical sound energy required to produce the same perceived loudness varies with frequency”.

Second, this frequency dependency is a function of overall level.

The first statement is not particularly bothersome. Your auditory system is well calibrated to that. A voice sounds normal because it sounds like what you are used to, not because it has “constant sound energy” or “constant perceived loudness” with frequency.

The second statement is much more trouble. It basically says that if you amplify an acoustic source (even if you do it perfectly), the perceived spectral balance will change. This is a well known effect, and most of our home entertainment systems have actually and “automatic loudness compensation” that changes the system voicing with overall level. We actually contemplated adding this to the Personalized Amplification System™ but after some soul searching we thought it would be too intrusive on the musician.

The main corrections are at very low levels, and in most practical live music settings, the effect is pretty minor.

As a rule-of-thumb guideline, turn the bass up a notch as you turn the volume down.

@@ Line 1: / Line 1: @@
-  quote: I'm familiar with the term 'crossover,' but not really with its meaning.
+  ; I'm familiar with the term 'crossover,' but not really with its meaning.
 ''Audio crossovers are a class of electronic filters designed specifically for use in audio applications, especially hi-fi. A commonly used dynamic loudspeaker driver is incapable of covering the entire audio spectrum all by itself. Thus, crossovers serve the purpose of splitting the audio signal into separate frequency bands which can be handled by individual loudspeaker drivers optimized for those bands. A combination of multiple drivers each catering to a different frequency band constitutes most hi-fi speaker systems. An audio crossover may also be constructed mechanically and is commonly found in full-range speakers.''
@@ Line 27: / Line 27: @@
 [[Bass Frequencies]]
+=== More Bass Talk ===
+Hilmar-at-Bose explained in  [http://bose.infopop.cc/eve/forums/a/tpc/f/3976055944/m/6121039062?r=2641067362#2641067362 More Bass Talk]
+====  Crossover ====
+If there is no B1 and nothing connected to the Bass Line Out. The L1 sees frequencies from 110Hz up. Feeding it anything lower, doesn’t make sense, since it couldn’t produce any acoustic output and if would rip the drivers to shreds.
+In any other case the L1 sees signals only from 180Hz up. There is no other variation in frequency or gain for the L1 no matter what else happens
+==== Bass Line Out and B1 behavior ====
+This is based on the design goal that “You should always sound the same; no matter how much Bass stuff is attached” I can try to explain my view of why this is a good design goal (of which you may disagree) but let’s look at the actual behavior first.
+===== Without Bass Line out =====
+xB1: 40Hz-180Hz, B1 specific EQ, some nominal gain that we call 0dB
+xB1: 40Hz-180Hz, B1 specific EQ, -6dB as compared to nominal
+===== With Bass Line Out =====
+xB1: 40-180 Hz, flat, roughly the same gain as 2 B1
+xB1: 40Hz-180Hz, B1 specific EQ, -6dB as compared to nominal
+xB1: 40Hz-180Hz, B1 specific EQ, -12dB as compared to nominal
+What this complex behavior does is the following. No matter if you attach 1, 2, or 4 B1s, you will get pretty much the same balance between all combined B1s and the L1s. It’s a little off for 3, 5, 6, 7 & 8 B1s, but still reasonably close.
+==== Frequency content of an acoustic guitar ====
+Oldghm, you did some really interesting experiments there. However, you have to be really careful when using an RTA. You can feed these things a pure sine wave at 80 Hz and by turning it up make the 63 Hz and even the 40Hz LED light up. They will be lower than the 80 Hz LED, but still come on. That does NOT mean, that the sine wave contains any other frequency than 80 Hz (it certainly doesn’t). It only means that the RTA has a pretty limited frequency resolution. The 63 Hz LED will respond best to 63 Hz signal but it’s in no way “blind” to 80 Hz signal.
+Thus being said, the actual frequency content is not easy to determine. All sounds that have a pitch are certainly constraint to 80 Hz and up (in standard tuning) and there isn’t actually too much energy at the fundamental. However, the “non-pitched” sounds like a hard string attack or whacking the top with your hand can very well have lower frequencies. Unfortunately, I don’t have any hard data on that, but we will measure that at some point.
+==== Equal loudness curves ====
+Here is the bunch
+http://hyperphysics.phy-astr.gsu.edu/hbase/sound/eqloud.html
+These curves tell us two things:
+First, the same physical sound energy produces different perceived loudness depending on frequency. You can turn that around into “The physical sound energy required to produce the same perceived loudness varies with frequency”.
+Second, this frequency dependency is a function of overall level.
+The first statement is not particularly bothersome. Your auditory system is well calibrated to that. A voice sounds normal because it sounds like what you are used to, not because it has “constant sound energy” or “constant perceived loudness” with frequency.
+The second statement is much more trouble. It basically says that if you amplify an acoustic source (even if you do it perfectly), the perceived spectral balance will change. This is a well known effect, and most of our home entertainment systems have actually and “automatic loudness compensation” that changes the system voicing with overall level. We actually contemplated adding this to the Personalized Amplification System™ but after some soul searching we thought it would be too intrusive on the musician.
+The main corrections are at very low levels, and in most practical live music settings, the effect is pretty minor.
+As a rule-of-thumb guideline, turn the bass up a notch as you turn the volume down.