ProSoundWeb Community
Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums => LAB Subwoofer Forum => Topic started by: Mark Wilkinson on June 27, 2016, 08:43:18 PM
-
Here's snip of a pair of Orbit Shifters indoors.
Close mic'ed, raw response except for a +4db boost at 47hz.
My question is: given the fairly steep natural rolloff below about 35hz, is a HP filter really necessary?
I've always used one, a BW4 @25...but not really sure why other than 'supposed to'....
Thx, Mark
-
In my opinion EVERY LOUDSPEAKER should use a HP filter-no matter what the response is.
The reason is that freq below when it can "easily work", the driver will tend to "flop around".
This can easily cause physical damage to the driver-even if the power is less than the ratings.
Even if the signal does not contain those freq, there are all sorts of "things" that can cause freq lower than intended to get to the driver.
Loose connections, people hot plugging something etc.
It is far better to be safe-rather than sorry.
-
Below port tuning or horn loading the driver is no longer damped by the enclosure meaning it operates as if it was out of the box hanging in free air and could be driven to it's mechanical limits with relatively little power, so if you want to keep your drivers functional you would be well advised to always have a low cut/high pass filter engaged.
-
Thx guys,
I get what you're saying and have always told myself the same considerations.
Please do know I'm only asking about this particular horn loaded sub, and for when it is properly powered/ limited.....
Looking at the trace comparisons below, blue is raw, and purple is with a BW24 HP @30hz and a LR48 LP @ 100hz.
Sure HP rolls off faster, but it doesn't seem to add that much protection, .....
.....and the protection comes at the expense of either a bunch more phase rotation or latency, whichever poison is chosen..(latency in this example, note the unchanged phase traces)
I'm interested in dropping the HP because the most credible literature I can find on phase audibility, says best sound comes from linear phase throughout except for system high-pass, which equates to sub high pass.
Supposedly, getting phase to follow sub magnitude roll-off is ideal. IOW if magnitude rolls off at 24 db/oct, phase should rotate at same 4 pole rate per octave.
Want to find out for myself...and at crankin volume ;D
-
Just because it can't produce 5hz, doesn't mean it ain't trying.
-
All good reasons. The Orbit Shifter(and any other sealed back/rear enclosure design) will offer some resistance to out of bandwidth response, but it is still a wise thing to have a good high pass filter in the circuit.
One other benefit is that it should help lower nonlinear distortion should the driver be in the situation of reproducing that 20hz organ pedal note along with the content within it's design capacity.
Best regards,
John
-
Here's snip of a pair of Orbit Shifters indoors.
Close mic'ed, raw response except for a +4db boost at 47hz.
My question is: given the fairly steep natural rolloff below about 35hz, is a HP filter really necessary?
I've always used one, a BW4 @25...but not really sure why other than 'supposed to'....
Thx, Mark
Yes. Use the Filter, young Skywalker.
-
Thx guys,
I get what you're saying and have always told myself the same considerations.
Please do know I'm only asking about this particular horn loaded sub, and for when it is properly powered/ limited.....
Looking at the trace comparisons below, blue is raw, and purple is with a BW24 HP @30hz and a LR48 LP @ 100hz.
Sure HP rolls off faster, but it doesn't seem to add that much protection, .....
.....and the protection comes at the expense of either a bunch more phase rotation or latency, whichever poison is chosen..(latency in this example, note the unchanged phase traces)
I'm interested in dropping the HP because the most credible literature I can find on phase audibility, says best sound comes from linear phase throughout except for system high-pass, which equates to sub high pass.
Supposedly, getting phase to follow sub magnitude roll-off is ideal. IOW if magnitude rolls off at 24 db/oct, phase should rotate at same 4 pole rate per octave.
Want to find out for myself...and at crankin volume ;D
The real answer is "it depends".
I do agree that the best sonic performace will be without a HP filter.
HOWEVER-if you want to push the system hard, the additional protection of the HP filter far outweighs any sonic advantage-ESPECIALLY when the driver STOPS WORKING because it has "beat itself to death".
I am sure that "all of the people in the audience" will "understand" that earlier they were enjoying a better sonic performance, and now there is no bass.
People who have been drinking are very understanding in that regard ;) ;) ;) ;)
I am sure that the club owner will not mind if people stop drinking and leave because there is no bass anymore.
After all, it sounded so good earlier, it more than makes up for the lack of bass now.
YES, I am being sarcastic, but also being very real.
Sometimes it is more important to keep the system working, than to get a "little bit more" out of it and then have it fail.
I remember back in the early 80s, working with a Metal band.
Everything was going great, but the band wanted a little bit more out of the system. They said they would pay for any blown speakers.
That may be all well and good, but at midnight on a Saturday-what am I going to do to get new drivers in the cabinets to play till 2am?
If I pushed the system and it stopped, they would QUICKLY forget all about the "we will pay for new speakers" and get mad at me for killing the system.
Then they and myself would not be hired back in the club because the system died.
None of the audience was complaining-just a couple of guys in the band.
It was simply not worth the risk.
Just something to consider.
I still say "ALWAYS use a properly set high pass filter". The possibility of damage is simply not worth it. Unless you are doing low to mid level listening.
-
Everything was going great, but the band wanted a little bit more out of the system. They said they would pay for any blown speakers.
Even if they survive the night, the damage may be done and not show itself for days, weeks, or months. Get cash payment up front as a non-refundable deposit.
-
Hey thanks all....I feel the experience/wisdom...HP in ;)
-
Won't you get more useable power from your amps if you add the hipass? Your box may not produce much below 40 hz but nobody told the amp that. It's still amplifying at least another octave below that.
-
Won't you get more useable power from your amps if you add the hipass? Your box may not produce much below 40 hz but nobody told the amp that. It's still amplifying at least another octave below that.
Especially if you're getting sub-sonic noise from somewhere in the system.
-
Won't you get more useable power from your amps if you add the hipass? Your box may not produce much below 40 hz but nobody told the amp that. It's still amplifying at least another octave below that.
Yes, and it will also clean up the sound in a different way.
Since the lower freq will cause the driver to move more than it should.
The more any loudspeaker moves, the more distortion it will produce.
So by keeping the low freq that cause excessive movement out of the driver-the other freq will be cleaner.
-
What's this new obsession that some folks have with content below 35Hz? Is it only in EDM? Because, frankly, I don't miss it in my music.
(This is not aimed at the OP, I'm just asking.)
-
What's this new obsession that some folks have with content below 35Hz? Is it only in EDM? Because, frankly, I don't miss it in my music.
(This is not aimed at the OP, I'm just asking.)
Hi Gordon, yeah, I don't listen to much below 35hz either.
From what little EDM I have heard, my guess is that below 35hz is more about feel, bodily vibration, than auditory.
Or maybe better said, integrating body vibration with auditory vibration.
Almost ridiculously, I've been a phase=o=holic lately.
My interest in removing a HP filter is not about extending LF, but about optimal phase alignment for highest possible sound quality.
Here's a post from a guy that makes alot of sense to me, and matches what many others are starting to say about linear phase.
http://www.diyaudio.com/forums/multi-way/283068-mini-synergy-horn-experiment-25.html#post4554771
I had been trying to flatten phase all the way down to lowest freq, but it wasn't sounding right, so now I'm trying to match magnitude and phase rolloff.
So, this HP filter question is really just about trying to follow the recommendation in the link regarding system highpass phase.
A HP filter in play makes doing it a bit trickier...
I wholeheartedly agree with linear phase improving transients, increasing vocal clarity, and giving bass impact.
The transient improvement is a little difficult to judge sometimes, as I've found I can be fooled by small amounts of HF/VHF change.
Bass can be tough too, with reflections and all.....
But vocal clarity never seems to lie...my trusted test is how well backing vocals come through.
Hey, if we ever get speakers that go to zero hz, then maybe flat linear phase to zero, with no system highpass.....
..... will fully integrate being bodily rocked and audibly rolled ;D
-
What's this new obsession that some folks have with content below 35Hz? Is it only in EDM? Because, frankly, I don't miss it in my music.
(This is not aimed at the OP, I'm just asking.)
There is A LOT of "normal" music that has content below 35Hz.
The low B on a bass guitar is 31Hz. Lots of orchestral bass drums are tuned very low. Many pipe organs go well below 35Hz-even down to 8Hz
Contra bass bassoons and clarinets can get to 15Hz
The low note on the piano is 27Hz Some pianos go down to 16Hz
Several brass instruments can get down well below 35Hz as well
Yes, I agree that these are not "normal rock instruments", but there is A LOT of music that is not rock that has instruments that go very low.
And not being able to reproduce the fundamental is not accurately reproducing their sound.
BTW-lots of synthesized pop music has content below 35Hz-but many people say you don't need it-simply because THEY have never heard it. But that doesn't mean it does not exist!
When doing a demo last year, I had a guy say about a "pop" song- "So THAT'S what that song is supposed to sound like"-after he heard the real extension.
YES-it DOES matter :)
Here is a link to help
http://www.contrabass.com/pages/frequency.html
-
Won't you get more useable power from your amps if you add the hipass? Your box may not produce much below 40 hz but nobody told the amp that. It's still amplifying at least another octave below that.
This is not quite right, because cutting the lower octave doens't add more useable power. The peak power is always the same, so the clipping voltage is also the same. When you cut out some portion of the spectrum, you only reduce the RMS power of the signal, but amplifier doens't care about it unless its power supply is very undersized for that amp.
-
...And not being able to reproduce the fundamental is not accurately reproducing their sound.
BTW-lots of synthesized pop music has content below 35Hz-but many people say you don't need it-simply because THEY have never heard it. But that doesn't mean it does not exist!
When doing a demo last year, I had a guy say about a "pop" song- "So THAT'S what that song is supposed to sound like"-after he heard the real extension.
YES-it DOES matter :)
Perhaps we still get a glimpse of what's there even when it is missing.
https://en.wikipedia.org/wiki/Missing_fundamental (https://en.wikipedia.org/wiki/Missing_fundamental)
-
This is not quite right, because cutting the lower octave doens't add more useable power. The peak power is always the same, so the clipping voltage is also the same. When you cut out some portion of the spectrum, you only reduce the RMS power of the signal, but amplifier doens't care about it unless its power supply is very undersized for that amp.
Average power matters at LF.
Higher frequency signals will be superimposed on top of the LF signals so can saturate or clip the amplifier sooner if they are not removed.
It is good practice to high pass the audio path appropriately for components bandwidth, while there should be some HPF already present in the audio chain.
JR
-
This is not quite right, because cutting the lower octave doens't add more useable power. The peak power is always the same, so the clipping voltage is also the same. When you cut out some portion of the spectrum, you only reduce the RMS power of the signal, but amplifier doens't care about it unless its power supply is very undersized for that amp.
ACTUALLY--------
By decreasing the bandwidth of a signal WILL reduce the power that it is putting out.
This is real easy to see for yourself.
Put in a noise source into an amp (no speaker hooked up) that is full bandwidth.
Hook a voltmeter to the output.
Now reduce a LP filter or raise a HP filter while looking at the voltmeter.
You will see the voltage lower as the bandwidth gets narrower.
This is because there are not as many freq adding together-causing the voltage applied to the load to be higher.
-
Now reduce a LP filter or raise a HP filter while looking at the voltmeter.
You will see the voltage lower as the bandwidth gets narrower.
This is because there are not as many freq adding together-causing the voltage applied to the load to be higher.
I'm talking about signals like white noise wich has a flat response over the band. In this case when cutting out, the RMS voltage will decrease as i said, but not the Peak voltage... (at least for low-mid order filters). Of course if you use a brick wall slope and filter out with a very narrow bandwidth, let's say one octave only (50-100 hz), the Peak voltage will also decrease because there are not as many freq adding together. But this will never be the case.
Different story is when you have to filter out to prevent from mic's dropping, or when you have a mix with extreme bass content wich has 30 hz +15 db respect to 100 hz. In this case the 30 hz foundamental leads the peak voltage, so filtering out it, will also decrease the peak voltage going to amplifier.
-
I'm talking about signals like white noise wich has a flat response over the band. In this case when cutting out, the RMS voltage will decrease as i said, but not the Peak voltage... (at least for low-mid order filters).
I'm sorry-but I disagree.
Anytime you have 2 signals of different freq, the peaks HAVE to be higher than either of the originals.
At some points.
Those points are where the signals are in phase (I am NOT talking about polarity here). They will "ride" on top of each other.
Of course at the points at which they are not in phase, the output voltage will be lower than either one.
So in order to reproduce both (or more) of the signals, the amp MUST have the voltage swing available to reproduce those peaks.
So the fewer freq that an amp has to reproduce-then the lower the possibility of the signals "stacking up" and increasing the maximum voltage (and resultant power).
-
ACTUALLY--------
By decreasing the bandwidth of a signal WILL reduce the power that it is putting out.
This is real easy to see for yourself.
Put in a noise source into an amp (no speaker hooked up) that is full bandwidth.
Hook a voltmeter to the output.
Now reduce a LP filter or raise a HP filter while looking at the voltmeter.
You will see the voltage lower as the bandwidth gets narrower.
This is because there are not as many freq adding together-causing the voltage applied to the load to be higher.
Does these analogie work?
A car has a certain amount of horse power available. That HP is available going up an incline. If the incline is "filtered out" does the HP increase? No, but top speed and acceleration increase.
A pickup with a bed full of gravel will struggle to pull a hill, but if the load is "filtered out" the hill is no problem.
In both cases, the transmission lasts longer and the engine sounds better while in operation.
-
Does these analogie work?
A car has a certain amount of horse power available. That HP is available going up an incline. If the incline is "filtered out" does the HP increase? No, but top speed and acceleration increase.
A pickup with a bed full of gravel will struggle to pull a hill, but if the load is "filtered out" the hill is no problem.
In both cases, the transmission lasts longer and the engine sounds better while in operation.
The technical term for this is superposition of waves, and X plus anything will be larger than X.
(https://encrypted-tbn1.gstatic.com/images?q=tbn:ANd9GcRzfGnTKGNwuUi7P5kM0zhlMlMGephWtgYmd3dcJfEsO16Q3ofipg)
For another analogy, imagine you are standing up in a row boat, in a river underneath a bridge and your head barely clears the bottom of the bridge. Just then a power boat races past creating a huge wake (the extra LF content). As the boat bobs up and down in the wave created by the power boat, your head crashes into the bottom of the bridge...
That is the physical manifestation of inadequate headroom (pun intended). ;D
JR
-
For another analogy, imagine you are standing up in a row boat, in a river underneath a bridge and your head barely clears the bottom of the bridge. Just then a power boat races past creating a huge wake (the extra LF content). As the boat bobs up and down in the wave created by the power boat, your head crashes into the bottom of the bridge...
That is the physical manifestation of inadequate headroom (pun intended). ;D
JR
I like that analogy. :)
-
The technical term for this is superposition of waves, and X plus anything will be larger than X.
JR
I will bow (pun intended) to you wave bob analogy, but if the "anything" added to X is a "negative anything" it will yield something smaller than X. ;)
-
The technical term for this is superposition of waves, and X plus anything will be larger than X.
(https://encrypted-tbn1.gstatic.com/images?q=tbn:ANd9GcRzfGnTKGNwuUi7P5kM0zhlMlMGephWtgYmd3dcJfEsO16Q3ofipg)
For another analogy, imagine you are standing up in a row boat, in a river underneath a bridge and your head barely clears the bottom of the bridge. Just then a power boat races past creating a huge wake (the extra LF content). As the boat bobs up and down in the wave created by the power boat, your head crashes into the bottom of the bridge...
That is the physical manifestation of inadequate headroom (pun intended). ;D
JR
Thinking about that analogy makes my head hurt.
-
I will bow (pun intended) to you wave bob analogy, but if the "anything" added to X is a "negative anything" it will yield something smaller than X. ;)
and i thought I was the pedantic one... :o
For polarity or phase to have any affect on combined waveforms they would have to be the exact same frequency.
In the context of a HPF the summed waveforms by definition will be different frequencies so combining will always result in higher peaks and more likely amplifier clipping.
QED
JR
-
Thinking about that analogy makes my head hurt.
Here is another "real world" example.
I have 2 sine waves-both of equal amplitude.
But when I turn them both on, the result get larger and smaller-depending on how the phase relationship of the waves is.
This was done on an old CRT scope-so that is why some of the traces are not "complete". You are just seeing what is being traced on the screen at the time the shutter on the phone is open.
But you should get the idea.
You can easily see how the higher freq wave is "riding" on the low freq wave and how the low freq wave is pushing the HF wave much higher in level (as in JRs example of a boat)
If the amplifier does not have enough voltage swing (power), it cannot reproduce the HF signal-resulting in distortion.
The top 2 photos are the individual freq. Both are clean
The bottom left is when they are added together-with plenty of headroom on the amp.
The bottom right is when each signal is just a little under clipping. So when they are added together, you end up with a greatly clipped signal.
If the lower freq signal was filtered out, then the higher freq one would still be nice and clean.
NOTE: I did readjust the gain on the scope channel to make it fit onto the screen.
In the first 3 photos, no gain was adjusted.
-
And to JR
I do realize that my scope was not centered, and the "clean addition" runs off scale.
But since this is not a "paper" I did not worry about doing a great presentation.
Hopefully the idea comes across-especially since this is NOT a simulation, but actual signals and how they add.
-
And to JR
I do realize that my scope was not centered, and the "clean addition" runs off scale.
But since this is not a "paper" I did not worry about doing a great presentation.
Hopefully the idea comes across-especially since this is NOT a simulation, but actual signals and how they add.
No worries good pictureds... if you think about it that is the exact mechanism behind IMD (intermodulation distortion), any level dependent non-linearity (like an amplifier clipping) will cause distortion products that are modulated by the interfering signals. A significant characteristic about IMD vs THD is that harmonic distortion products are are always up at higher multiples, while IMD is routinely LF mud and grunge.
JR
-
Hopefully not too much of a swerve but what do you call the acoustic modulation of a higher frequency from a driver that is also producing a lower frequency? As in the physical manifestation of Ivan's scope traces? I've always thought of this as IMD but I saw some recent threads here that said it wasn't. That old gargley sound we got when we tried to plug a mic and sing though a spare channel on the bass amp.
-
Hopefully not too much of a swerve but what do you call the acoustic modulation of a higher frequency from a driver that is also producing a lower frequency? As in the physical manifestation of Ivan's scope traces? I've always thought of this as IMD but I saw some recent threads here that said it wasn't. That old gargley sound we got when we tried to plug a mic and sing though a spare channel on the bass amp.
A lot of that "sound" was simply because the bass amp did not have any higher freq components to reproduce the higher freq.
The typically larger speakers used in bass amps have an inherent HF roll off due to the self inductance of the voice coils-that acts as a low pass filter.
Also a lot of bass amps have a very low damping factor-which is part of the "sound".
So when the speakers try to "return to zero", they overshoot the mark and ten d flop around.
The vocal freq are just "along for the ride".
Consider also there could be a bit of the doppler effect on bass cones that are moving a good distance that would further screw up the sound.
The cone movement is pretty much like the sine waves I showed. The lower freq is moving it in and out-with the higher freq vibrating it.
It is also like the ocean. There are basically 2 effects on the water. The tide is like low freq, producing large rolling waves, while the wind is like the high freq producing ripples on the water.
When the wind is blowing, there are ripples riding on top of the large tidal waves.
Also think about what it takes to stop those waves.
A pier post will have a smooth area behind it where the size of the post is large compared to the ripples the wind created.
But that post is small as compared to the size of the tidal wave. It takes a MUCH LARGER object to stop the action of the tidal wave.
This needs to be kept in mind when considering acoustical treatment-size of horns-spacing of elements and so forth.
-
Hopefully not too much of a swerve but what do you call the acoustic modulation of a higher frequency from a driver that is also producing a lower frequency? As in the physical manifestation of Ivan's scope traces? I've always thought of this as IMD but I saw some recent threads here that said it wasn't. That old gargley sound we got when we tried to plug a mic and sing though a spare channel on the bass amp.
I'm not the speaker expert but I vaguely recall a mechanism called "Doppler distortion" a variation on the effect that shifts the pitch of the train whistle between approaching and leaving...
IMD involves a nonlinearity, which could occur in drivers at high amounts of excursion.
Yes we are veering a bit, but such things happen on the internets. 8)
JR
-
Thanks to all for the helpful replies...made me think...
I liked the boat analogy and o-scope shots ... they seem to strongly support the case for linear-phase throughout the spectrum.
I mean, I'm going to hit my head harder when in the boat, or go off the scope screen further, when waves are peak summing / exactly in phase.
Also I read that radio broadcasters have used all pass filters as phase rotators to help facilitate transmission..phase rotation being a means of signal compression that was needed to lower transmission peaks.
Several of you mentioned non linear distortion increases that could occur from a driver being pushed below tuning...
Makes sense, and made me realize I'm kidding myself about listening for 'pure sound quality' at crankin volume.
If I want to listen for phase effects, it only makes sense it has to be at levels before THD, IM, or whatever, starts to rise with increased output.
I guess SQ at cranking volume comes down to enough rig for the gig, at least as much as it does for proper tuning and gear quality...
Which begs a topic swerve if I may.... how do you assess when your rig's SQ is falling off with volume?
What particularly do you listen for?
Or for what, and how do you measure the 'for what'? (i'm assuming the 'for what' is some form of distortion...)
Not asking about when one band running out of steam before another, like subs crapping out before mains, etc...
But just about when system linearity starts falling apart.....
-
Thanks to all for the helpful replies...made me think...
I liked the boat analogy and o-scope shots ... they seem to strongly support the case for linear-phase throughout the spectrum.
I mean, I'm going to hit my head harder when in the boat, or go off the scope screen further, when waves are peak summing / exactly in phase.
The actual phase will make only a VERY little difference-and just in a different way-not better or worse.
If you have 2 freq that are different-whether or not they are "in phase", they will add and subtract just as much.
Just a different places along a "absolute phase" line.
The amp does not care what the phase is-only the maximum voltage that is going to it.
And that will remain the same-even if the phase of one of the signals is changes.
I DO AGREE that you can set up a test that for freq that are very close in freq, and "misadjust" the phase, for some periods of time the output will be lower. But for OTHER TIMES, it will be just as high.
The maximum and minimum values would not change-they would just be in different places
-
The actual phase will make only a VERY little difference-and just in a different way-not better or worse.
If you have 2 freq that are different-whether or not they are "in phase", they will add and subtract just as much.
Just a different places along a "absolute phase" line.
The amp does not care what the phase is-only the maximum voltage that is going to it.
And that will remain the same-even if the phase of one of the signals is changes.
I DO AGREE that you can set up a test that for freq that are very close in freq, and "misadjust" the phase, for some periods of time the output will be lower. But for OTHER TIMES, it will be just as high.
The maximum and minimum values would not change-they would just be in different places
Agreed. I wasn't thinking so much about amp loads.
By 'in phase', I mean linear phase throughout the spectrum that allows for perfect timing and summing of each the contributing musical waveforms.
IOW, a flat phase trace in smaart.
So that a drum whack, whacks harder and cleaner...having all its waveform components sum and peak perfectly.
I think this is what linear phase throughout the spectrum does, by nailing timing.
IME, it adds alot to the realism of transients, and overall clarity.
I think Meyer's long standing design goal of a mildly down-sloping phase trace has been spot on, well, at least until FIR arrived 'with no slope needed'.
-
Agreed. I wasn't thinking so much about amp loads.
By 'in phase', I mean linear phase throughout the spectrum that allows for perfect timing and summing of each the contributing musical waveforms.
IOW, a flat phase trace in smaart.
So that a drum whack, whacks harder and cleaner...having all its waveform components sum and peak perfectly.
I think this is what linear phase throughout the spectrum does, by nailing timing.
IME, it adds alot to the realism of transients, and overall clarity.
I think Meyer's long standing design goal of a mildly down-sloping phase trace has been spot on, well, at least until FIR arrived 'with no slope needed'.
The discussion-and my responses have been totally about the effects of using a HP filter to reduce the overall voltage going to and coming of an amp.
The freq will add-no matter what the phase is.
The response of a system is a completely different issue.
-
The discussion-and my responses have been totally about the effects of using a HP filter to reduce the overall voltage going to and coming of an amp.
The freq will add-no matter what the phase is.
The response of a system is a completely different issue.
Hi Ivan, .....well I need to disagree here a little :)
....as OP, all of my posts have been about optimizing system response, initially asking if dropping HP to do so was foolish.......then moving to asking about how to gauge non-linear distortion (something I would still love to hear about...maybe a new thread...)
And yes, freq will add no-matter what the phase is.
But isn't the timing of when those frequencies are produced/added, dependent on phase???
IME, phase-included timing determines transient, or rather impulse, response,..... which surely is system response. ;)
-
Hi Ivan, .....well I need to disagree here a little :)
....as OP, all of my posts have been about optimizing system response, initially asking if dropping HP to do so was foolish.......then moving to asking about how to gauge non-linear distortion (something I would still love to hear about...maybe a new thread...)
And yes, freq will add no-matter what the phase is.
But isn't the timing of when those frequencies are produced/added, dependent on phase???
IME, phase-included timing determines transient, or rather impulse, response,..... which surely is system response. ;)
As usual-it depends.
If you are talking about lower to mid level usage-then going without a HP filter WILL have a better phase response, can add a bit more depth to the sound and so forth-all good things.
HOWEVER-if the system is being pushed anywhere near capability (of the amp or the speaker) then adding HP filter can help to improve the longevity of the speaker and reduce the clipping of the amplifier. Both good things.
Regarding the phase-if you look over a longer period of time (not just a couple of cycles) you will see that when you have 2 different freq they will add together just as often when they are "inphase" as when they are "out of phase".
This would not matter whether they are 10* or 180* out.
And how would you determine whether they are in or out of phase? How much difference is acceptable for "in phase".
Remember that the phase angle only relates to some other point in time.
You can take a sine wave and change the phase all day long and you will never be able to hear it-UNLESS you hear it during the actual change.
If it is stopped between each change-there is no way you can say if the signal is in or out of phase or even polarity or if you "shifted" it by 100,000 degrees.
It is all REFERENCED to something else.
If that other thing is missing-there is no reference.
What if I asked you "I am thinking about adding another 1,000 watts of power to my system-will it get louder?"
If my current system is a couple of hundred watts-sure.
But if my system is 100,000 watts, nobody would notice.
-
As usual-it depends.
If you are talking about lower to mid level usage-then going without a HP filter WILL have a better phase response, can add a bit more depth to the sound and so forth-all good things.
HOWEVER-if the system is being pushed anywhere near capability (of the amp or the speaker) then adding HP filter can help to improve the longevity of the speaker and reduce the clipping of the amplifier. Both good things.
Sure, all this makes good sense.
Regarding the phase-if you look over a longer period of time (not just a couple of cycles) you will see that when you have 2 different freq they will add together just as often when they are "inphase" as when they are "out of phase".
This would not matter whether they are 10* or 180* out.
And how would you determine whether they are in or out of phase? How much difference is acceptable for "in phase".
Remember that the phase angle only relates to some other point in time.
You can take a sine wave and change the phase all day long and you will never be able to hear it-UNLESS you hear it during the actual change.
If it is stopped between each change-there is no way you can say if the signal is in or out of phase or even polarity or if you "shifted" it by 100,000 degrees.
It is all REFERENCED to something else.
If that other thing is missing-there is no reference.
This too makes sense. However, I don't really think comparing sine waves of different freqs has much value, other than a starting point in seeing how waveforms sum electrically.
Like you say, what's the reference?
I'd say the reference is the system input.
And that phase is determined by comparing system output to system input, freq by freq.....IOW, a plain ole smaart or similarly derived phase trace.
So to me, flat phase or linear phase, just means input and output are in phase, so that the system isn't outputting mistimed waveform pieces that are supposed to arrive together.
And that this is a real key to clean impulse response.
-
Sure, all this makes good sense.
This too makes sense. However, I don't really think comparing sine waves of different freqs has much value, other than a starting point in seeing how waveforms sum electrically.
Like you say, what's the reference?
I'd say the reference is the system input.
And that phase is determined by comparing system output to system input, freq by freq.....IOW, a plain ole smaart or similarly derived phase trace.
So to me, flat phase or linear phase, just means input and output are in phase, so that the system isn't outputting mistimed waveform pieces that are supposed to arrive together.
And that this is a real key to clean impulse response.
Phase MUST have a freq associated with it.
If somebody says something is "out of phase"-what does that mean? how much at what freq?
It is all relative.
With any processor as soon as you start to add xovers or eq, you are out of phase at some point/freq
-
This too makes sense. However, I don't really think comparing sine waves of different freqs has much value, other than a starting point in seeing how waveforms sum electrically.
The reason for showing 2 sine waves was that the discussion at the time was about signals adding together and hitting the maximum output (voltage swing) of the amplifier.
This can be shown with music-but it is nowhere nearly as clean or easy to see for those not accustomed to looking at such things.
The 2 sine waves are easy to demonstrate this with.
Phase has nothing to do with this particular part of the discussion.
-
Phase MUST have a freq associated with it.
If somebody says something is "out of phase"-what does that mean? how much at what freq?
It is all relative.
With any processor as soon as you start to add xovers or eq, you are out of phase at some point/freq
Hi Ivan, please pardon my inability to convey sometimes....
What I am trying to say is, instead of thinking about "out of phase" as one freq vs another, how about thinking about it terms of input to output....again, like a dual FFT phase trace.
So "out of phase" then simply becomes a phase shift for any single given freq, comparing the frequency's phase at input vs its phase at output.....
FIR processors with linear-phase xovers and eqs can keep it all "in phase", but of course at the expense of latency.
BTW, Merlijn just posted a killer good video about sub impulse response in the Measurement forum.
It gets at all I've been trying to test for and listen to.....
-
Hi Ivan, please pardon my inability to convey sometimes....
What I am trying to say is, instead of thinking about "out of phase" as one freq vs another, how about thinking about it terms of input to output....again, like a dual FFT phase trace.
So "out of phase" then simply becomes a phase shift for any single given freq, comparing the frequency's phase at input vs its phase at output.....
FIR processors with linear-phase xovers and eqs can keep it all "in phase", but of course at the expense of latency.
BTW, Merlijn just posted a killer good video about sub impulse response in the Measurement forum.
It gets at all I've been trying to test for and listen to.....
But that has nothing to do with the headroom capability of an amp and what happens when you reduce low freq information that is not needed.
-
But that has nothing to do with the headroom capability of an amp and what happens when you reduce low freq information that is not needed.
Hey Ivan, no offense but I'd have to be a moron not to realize that....
Trying to keep the thread on original ground....where you coming from?
-
Hey Ivan, no offense but I'd have to be a moron not to realize that....
Trying to keep the thread on original ground....where you coming from?
I was simply talking about the posts that were regarding the extra low freq energy "gobbling up" the amp headroom.
Not to be "mean", but you were the one that brought up the while idea of phase.
And all I was trying to say is that phase has nothing to do with amplifier headroom when signals are adding together.
Maybe I am missing something
-
Hey Ivan, no offense but I'd have to be a moron not to realize that....
Trying to keep the thread on original ground....where you coming from?
Ivan is on topic.
This thread was finished some time ago.
JR
-
Ivan is on topic.
This thread was finished some time ago.
JR
About 1 page back...
-
Did somebody say phase? Wait, I've heard of that.
When two frequencies collide, what's out of phase here is in phase there. And maybe a few degrees less over there.
Gimmie a beat!
Phasers set to stun captain.
-
Did somebody say phase? Wait, I've heard of that.
When two frequencies collide, what's out of phase here is in phase there. And maybe a few degrees less over there.
Gimmie a beat!
Phasers set to stun captain.
:) :)
-
To answer the original question, I'd say you absolutely need a HP filter.
I recently destroyed a Beyma 15P1200Nd during testing, as the software was running the sweeps an octave past what I asked it to do (turns out its in the manual, who knew?). Set for 30Hz-100Hz sweep. Cabinet is a 1/4-wave transmission line that tunes to the mid-30s. Push the levels up for a high-power sweep and bam, 500w+ of 15Hz sine. The driver is rated for (and manages) 52mm p/p travel before damage, which was very impressive to watch. The cone got itself some fold lines and had to be reconed, to the tune of over £100.
Sure, the phase will go wonky at the bottom of the passband, adding a little group delay (the audibility of which is contested). Better than your bass going to mush because the kick drum port puts out a poof of air at infrasonic frequencies and you haven't set a simple filter. Or maybe someone will drop a mic. Whatever. Its not worth the risk.
Chris
-
Sure, the phase will go wonky at the bottom of the passband, adding a little group delay (the audibility of which is contested). Better than your bass going to mush because the kick drum port puts out a poof of air at infrasonic frequencies and you haven't set a simple filter. Or maybe someone will drop a mic. Whatever. Its not worth the risk.
Chris
It is all mater of compromise. It is real easy to get "hung up" on the little "perfections" and then destroy the loudspeaker.
Often having the speaker live to play another day is more important than somebody being happy with a proper "alignment".
-
Below port tuning or horn loading the driver is no longer damped by the enclosure meaning it operates as if it was out of the box hanging in free air and could be driven to it's mechanical limits with relatively little power, so if you want to keep your drivers functional you would be well advised to always have a low cut/high pass filter engaged.
Not entirely true. The Orbit Shifter has a sealed rear volume for the driver, so there is always some physical resistance to cone motion below resonance.
Best regards,
John
-
Not entirely true. The Orbit Shifter has a sealed rear volume for the driver, so there is always some physical resistance to cone motion below resonance.
Best regards,
John
John,
The operative word is "some", which is the same air spring as expected from a small rear chamber, same as any sealed sub. If you measure it's distortion below Fc, you will notice the excursion rapidly exceeds Xmax with far less than full power, though not quite as badly as a TH or BR below Fb.
Art
-
That may be all well and good, but at midnight on a Saturday-what am I going to do to get new drivers in the cabinets to play till 2am?
IIRC the late, great M.L. Procise once said: "BAD sound beats NO sound every time"