Sure, but not can result, ...it will result.....if xovers are passive or active IIR.
Imho, the significant group delay from using steep passive or IIR active xovers, is why so many people shun their sound, and prefer low order.....
....Mr Dunlavy being one of them. I bet he would have loved to play with FIR and linear phase xovers.
'ain't that the truth. Although if I see one more company marketing spiel including the phrase 'designed without compromise'...
I had a 1000 bucks to spend. I spent it. I didn't compromise
There is a difference.
Have a look at my thread for a 4 way active here. Large 4 way using Active Crossovers
The Ultimate Preamp allows me to put whatever crossover I want into it. It allows me to compare apples for apples. Same drivers different xover. I use 1st order just like Mr Dunlavy did for the same reason. When I go to 2nd, 3rd,nth order, you can clearly hear difference. I think this is one of the hardest things to hear since no manufacture makes a passive speaker using the same drivers with different order xovers, so it does make it hard to compare because of how different all these speakers, rooms etc sound. I have done this test many times with friends, had them listen to the same song using different order xovers. They all agreed there is a difference and all had the same opinion of the overall sound. The 1st order sounds better. I can now put a song on when they come over and they can immediately tell I'm messing with their heads and tell me to change it back.
Unless you can compare without delay the same song in the same room on the exact same pair of loudspeakers, its very difficult. But their is a difference.
Everyone who I do this to say about the 1st order, the sound is more enveloping, it surrounds your head more. To me that is sound stage. They all agree as you go higher order the characteristics of each the individual drivers come through. You are no longer listening to the music you are now listening to the drivers, the speakers lose their blendedness, if their is such a word. You really have to hear for yourself.
JD did everything right. If you want to get rid of your passive xover, go active, JD would be happy to see you do that.
I had a 1000 bucks to spend. I spent it. I didn't compromise
Right, that's it, I'm getting the blunderbuss.
Re the above -I don't doubt you've found that. I am simply stating a fact however: not everybody is that fussed about linear-phase / transient perfect designs, myself included. I simply don't hear the huge advantages that are sometimes touted for them, leading me to the conclusion that these are often overstated and my own priorities lie elsewhere. Which doesn't change the fact that the Dunlavy designs were excellent (if a mite demanding on the tweeter at higher levels
).Attachments
Last edited:
Not disagreeing with you one bit Scott. I just wish everybody could come and listen to mine and see what difference time alignment and crossover order make to the sound. Being able to change any parameter instantly and hear the result is something we could have only dreamed about years ago. The human ear is an amazing thing. You wouldn't think the ear could discern 10uSec. When I remove the time alignment and then put it back 10uSec is clearly audible.
Thats the beauty of the active preamp. No distortion or blown tweeters no matter how high you turn the volume up. Johns Magnus would have been a reality by now if he were still here.
Its a testament to his brilliance that 15 years after the last pair of DAL rolled off the line that they are still so revered. I know he would be thrilled that we still talk about his work and his life.
Thats the beauty of the active preamp. No distortion or blown tweeters no matter how high you turn the volume up. Johns Magnus would have been a reality by now if he were still here.
Its a testament to his brilliance that 15 years after the last pair of DAL rolled off the line that they are still so revered. I know he would be thrilled that we still talk about his work and his life.
Last edited:
Perhaps. I'd also consider him capable of not requiring them, in the context of the kind of speakers he prefers. In addition could a steep filter cause an abrupt transition when bringing together two ways that are not well matched, such as to cause bad sound that a more shallow filter might reduce?
I had a 1000 bucks to spend. I spent it. I didn't compromise
just to clarify my speakers cost a hell of a lot more than 1000 to build. Add another zero.
I can't see how the type speakers he designed would benefit significantly with FIR either. Other than perhaps making timing a little easier, and opening up a wider selection of usable drivers.
Because his designs depend on well behaved wide-bandwidth drivers share the SPL load....as does any design trying to use first order xovers...
I think the only real limitation with such first order designs is in SPL and dynamics.
(As a rule imo, the wider the bandwidth of a driver, the lower its SPL potential.)
Dunlavy achieves SPL and dynamics with very large speakers,
using multi-way increasingly sized drivers from center, that continuously sum towards 1/4 wave length spacing.
But to maintain first order, the drivers must all have wide bandwidth relative to their primary range where they are the sole frequency producer.
And SPL depends on drivers sharing the load.
Steep filters in his designs would end up decreasing available SPL and dynamics.
Now, take the exact same physical box, and put more robust drivers of the same sizes in place.
Where all the drivers have narrower bandwidth and higher SPL and dynamic capability.
Then use steep linear phase filters to tie it together, and I'll wager you get an even better sounding, certainly more powerful speaker with greater headroom.
Oh, I don't understand your question ...
why bother theorizing over a two-way when drivers are not well matched?
Maybe you mean a shallow xover can hide sin better in such cases? I can see that...
Does a first order crossover ever equate to 6dB rolloff? Very rare IMO. Usually nearer LR2 and 12dB slopes. But in a three-way you might achieve something close to that. But fascinating to see the crossover at last.
And that time-alignment issue is a pain.
I am ready to pull the trigger on an LR2 design by Michael Chua: Finch (Morel CAT378 + Vifa PL18W0-09-08) – AmpsLab
Clever Morel CAT378 waveguide stuff that solves some dispersion issues along with time alignment on a flat baffle: MOREL CAT 378 Soft Dome Horn Tweeter – AmpsLab
Here in a DeVore Orangutan '96:
It is quite easy to use a simulator to look back down the filter with the inputs shorted. Just replace the driver with an amp. Seems like crossovers DO get in the way, compared to digitally filtered amps. They have high impedance at points near crossover.
Michael has quite won me over to simple filters, if the drivers are well-behaved enough. They do have a nice liveliness.
And that time-alignment issue is a pain.
I am ready to pull the trigger on an LR2 design by Michael Chua: Finch (Morel CAT378 + Vifa PL18W0-09-08) – AmpsLab
Clever Morel CAT378 waveguide stuff that solves some dispersion issues along with time alignment on a flat baffle: MOREL CAT 378 Soft Dome Horn Tweeter – AmpsLab
Here in a DeVore Orangutan '96:
It is quite easy to use a simulator to look back down the filter with the inputs shorted. Just replace the driver with an amp. Seems like crossovers DO get in the way, compared to digitally filtered amps. They have high impedance at points near crossover.
Michael has quite won me over to simple filters, if the drivers are well-behaved enough. They do have a nice liveliness.
I met John it in Utah where he has set up ship and has an opportunity to work with him and hung around for a few days to check it it out. John was cool but some of the technicians seemed off and kind of back stabby, so ultimately I passed in the job and went in to the job I had in pocket at Boeing.
After John’s company folded the first time when he first moved back to the states in the late 80s a group of us bought the entire factory inventory, which at the time were making Black Knights. I think we made around 35 pairs and blew out the rest of the inventory. At that time both the black knights and the sovereigns were using Dynaudio drivers. They were first rate. After that was closed out I used 4 of the 30W100s in a flour/ceiling sub tube coupled with an Esotar tweeter and two 15W75s in a D’Appolito satellite. Those were the days. I’m still using hardware, screws, registers and bits and bobs after all these years!!
My two cents would e to put new surround’s in the Vifa if possible because it’s just way less hassle and risk. I agree with others I wouldn’t do anything ‘big’ with them — just get them up and running.
After John’s company folded the first time when he first moved back to the states in the late 80s a group of us bought the entire factory inventory, which at the time were making Black Knights. I think we made around 35 pairs and blew out the rest of the inventory. At that time both the black knights and the sovereigns were using Dynaudio drivers. They were first rate. After that was closed out I used 4 of the 30W100s in a flour/ceiling sub tube coupled with an Esotar tweeter and two 15W75s in a D’Appolito satellite. Those were the days. I’m still using hardware, screws, registers and bits and bobs after all these years!!
My two cents would e to put new surround’s in the Vifa if possible because it’s just way less hassle and risk. I agree with others I wouldn’t do anything ‘big’ with them — just get them up and running.
It is probably not even necessary to be able to reproduce a perfect square-wave for a speaker to be subjectively accurate.
I am convinced that there is a certain window within which the group delay distortion has to be for not being audible. Unfortunately there are no generally agreed figures for that yet. One of the more recent and also quite interesting papers dealing with this came from Finland.
Regards
Charles
I am convinced that there is a certain window within which the group delay distortion has to be for not being audible. Unfortunately there are no generally agreed figures for that yet. One of the more recent and also quite interesting papers dealing with this came from Finland.
Regards
Charles
No argument here; I can't honestly say a 'perfect' square-wave response is a major priority at this end either. Very nice if you can get it and don't pay an excessive price in (arguably more important) areas to do so of course, so I'm not knocking it in a general sense.
GD is an interesting one though. I've sometimes used LR6 & LR8 acoustic in 2-ways, albeit with the [major] caveat that I like low XO frequencies, and since the audibility threshold varies with frequency that's likely why it worked for me in those systems. Either that or I'm just a tone-deaf monster. I usually try to stick within Blauert & Laws's approximate thresholds, so I wouldn't be over-inclined to use them above 1.5KHz or so. Below that -YMMV as always.
GD is an interesting one though. I've sometimes used LR6 & LR8 acoustic in 2-ways, albeit with the [major] caveat that I like low XO frequencies, and since the audibility threshold varies with frequency that's likely why it worked for me in those systems. Either that or I'm just a tone-deaf monster.
I usually try to stick within Blauert & Laws's approximate thresholds, so I wouldn't be over-inclined to use them above 1.5KHz or so. Below that -YMMV as always.
Last edited:
i should be thrown into the tone deaf crowd too
My take on square waves is that they simply reflect magnitude and phase across the spectrum. Totally flat mag and phase would produce perfect squares waves across the spectrum, other than the need for many harmonics above the audio frequency.
(If anyone cares, here's a good paper on square waves i recently linked on another thread Square Waves And DC Content: Deconstructing Complex Waveforms - ProSoundWeb)
So for me, when asking if square waves matter, it's really just a question of how important is flat phase,
since i believe we all basically agree that flat mag matters. (By flat mag, I mean before any house curve preference is overlay-ed.)
I think phase matters and find it very interesting, because intuitively it seems it has a number of potentially audible aspects.
There's the often discussed group delay.... how much at what frequency becomes audible.
And then there's timbre, which seems it must change as phase relationships change between the summing spectral components.
And then there's transients' individual amplitudes, where the relative phase timings produce higher or lower peaks as voltages sum via their phase angles.
In addition to higher order xovers mucking things up, we all know you can't lean on high-Q EQ's.
So yeah, I'm a phase-aholic....and love to see good square waves
But happily mainly due to listening and getting great measurements, rather than the intuitive speculations above ....
My take on square waves is that they simply reflect magnitude and phase across the spectrum. Totally flat mag and phase would produce perfect squares waves across the spectrum, other than the need for many harmonics above the audio frequency.
(If anyone cares, here's a good paper on square waves i recently linked on another thread Square Waves And DC Content: Deconstructing Complex Waveforms - ProSoundWeb)
So for me, when asking if square waves matter, it's really just a question of how important is flat phase,
since i believe we all basically agree that flat mag matters. (By flat mag, I mean before any house curve preference is overlay-ed.)
I think phase matters and find it very interesting, because intuitively it seems it has a number of potentially audible aspects.
There's the often discussed group delay.... how much at what frequency becomes audible.
And then there's timbre, which seems it must change as phase relationships change between the summing spectral components.
And then there's transients' individual amplitudes, where the relative phase timings produce higher or lower peaks as voltages sum via their phase angles.
In addition to higher order xovers mucking things up, we all know you can't lean on high-Q EQ's.
So yeah, I'm a phase-aholic....and love to see good square waves
But happily mainly due to listening and getting great measurements, rather than the intuitive speculations above ....