So when crossing a "supertweeter" (or just a very high-extension tweeter, no matter the nomenclature) over with a full-range driver with a single low value cap to really flesh out the topmost octave, is phase aligning them important at those nearing 20kHz and above frequencies?
I mean, as critical as phase alignment can be with a range of frequencies.
If I use capacitors as 1st order for the full-range anyway, are the drivers now phase coherent in terms of at least the signal itself? Would they not be if the full-range was ran entirely crossover-free?
In relation to this, how important is physical time-alignment?
I mean, as critical as phase alignment can be with a range of frequencies.
If I use capacitors as 1st order for the full-range anyway, are the drivers now phase coherent in terms of at least the signal itself? Would they not be if the full-range was ran entirely crossover-free?
In relation to this, how important is physical time-alignment?
This may be more of a power situation, and while phase can have a hand in the power distribution, it's not quite the same as regular setting of the phase shift. I say this because if you listened to a fullrange in an anechoic chamber dead on axis with all the EQ you needed, then you'd already be OK for 20kHz.
We are less sensitive to phase itself at these higher frequencies. Also consider that the fullrange may be heavily into lobing at these frequencies which may set the scene, so to speak.
We are less sensitive to phase itself at these higher frequencies. Also consider that the fullrange may be heavily into lobing at these frequencies which may set the scene, so to speak.
I have replaced Audio Physic Tempo V midranges with 4" Mark Audio full-ranges, the original tweeter (well over 20kHz capable) is with just 1uF (-6dB at 20kHz) and seems to integrate pretty well giving a very needed fill-in for near 20kHz transients. And naturally, less strict of listening position for soundstaging now.
The FR is crossed with 100+33+15+3,3 uF to the bass drivers.
The cabinet has an approx 7° slant so the tweeter is physically at about the same axis as the voice coil of the full-range beneath it.
The FR is crossed with 100+33+15+3,3 uF to the bass drivers.
The cabinet has an approx 7° slant so the tweeter is physically at about the same axis as the voice coil of the full-range beneath it.
About this specifically, does lobing become less likely with frequencies approaching 20kHz?
Lobing becomes more complex as the wavelengths grow smaller than the diaphragm.
The big question you have to ask yourself here is..
Am I going to cross in the regular way, considering directivity and all else that goes into a standard smooth cross? or..
Am I going to fill in the reduced power at the top end in the regular supertweeter way, varying level and frequency until I find the amount I want?
In addition to the supertweeter method however, you also have phase, polarity and the woofer upper cross as variables, to use as tools to refine your results. Now, unless you do complex analysis to begin with then I might first want to listen for problems before deciding that there is a need to fix something.
The big question you have to ask yourself here is..
Am I going to cross in the regular way, considering directivity and all else that goes into a standard smooth cross? or..
Am I going to fill in the reduced power at the top end in the regular supertweeter way, varying level and frequency until I find the amount I want?
In addition to the supertweeter method however, you also have phase, polarity and the woofer upper cross as variables, to use as tools to refine your results. Now, unless you do complex analysis to begin with then I might first want to listen for problems before deciding that there is a need to fix something.
Any BW above ~the VC diameter is dustcap modes, so technically yes if not a coax. FWIW/YMMV, etc., many of us 'hornies' (vintage at least) use 1.0 uF to fill in above ~10 kHz with no noticeable lobing, timing issues beyond hooking it up with the polarity that either sounds best, otherwise in positive.
Would love to have complex analysis skills in check for that but yes, there are no problems to my ears but then again I don't have an entirely phase-coherent and physically time aligned reference like a Maggie would be.
I plan on getting the Magnepan LRS just to have these benefits as a reference for any and all phase/time-alignment/polarity issues to deal with in other speakers.
The best reference is what sounds natural. Granted, that may not be easy with some speakers.. sometimes because of the way they have been intentionally designed, and sometimes due to flaws.
There are times when setting something up and tweaking it by ear can get you good results, and times when it is impossible to do by ear. Maybe this is one that's worth trying..?
There are times when setting something up and tweaking it by ear can get you good results, and times when it is impossible to do by ear. Maybe this is one that's worth trying..?
My personal experience many years ago in the vintage realm of helping out 2 and 3” horns with above 10khz were ribbons tilted downwards slightly. The narrow vertical radiation i found to be less destructive. Simple cap and Lpad to dial things in. Once I passed 40 though where above 14khz no longer mattered, it wasn’t worth the effort with any of my designs….I just don’t concern myself with response that high. If you’re in that camp, I suggest you take stock and act accordingly.
On a side note, we cue in around 8-9khz for height detection so in lies the problem….it’s the design and shape of our outer ear that reflects the sound for time domain comparison…..an evolutionary tool for hunting and self preservation. I began having to study this stuff extensively as I began mixing for ATMOS. For a stereo field, the problem for the listener and the design becomes that anomolies in the time domain here can crush imaging as intended by the recording. With this recent understanding of the importance of time domain and biology, I make sure that none of my designs phase overlap in this area…….and when mixing, I put a small 1.5 db bump or cut with a high Q centered on 8k on voices and time domain busses such as reverb and delay…..also often with electronic/synth sounds and effects that are meant for ambience in the recording.
Next time you’re listening, take your fore finger and ever so slightly tilt your outer ear away from your head where it meets your skull……it will come as quite a shock how massive and immediate the changes to the aural experience. This has been my argument for decades when we talk about measurements and what we know and what we don’t…..audiophilia becomes more relevant and the science of speaker design less so as there are two huge variables in the listening equation…..the 3d space and the biology of the listener…..notice how I didn’t use capability as to exclude those without the ‘skill’…..that’s where audiophoolery and the music industry went horribly wrong by not understanding their market.
On a side note, we cue in around 8-9khz for height detection so in lies the problem….it’s the design and shape of our outer ear that reflects the sound for time domain comparison…..an evolutionary tool for hunting and self preservation. I began having to study this stuff extensively as I began mixing for ATMOS. For a stereo field, the problem for the listener and the design becomes that anomolies in the time domain here can crush imaging as intended by the recording. With this recent understanding of the importance of time domain and biology, I make sure that none of my designs phase overlap in this area…….and when mixing, I put a small 1.5 db bump or cut with a high Q centered on 8k on voices and time domain busses such as reverb and delay…..also often with electronic/synth sounds and effects that are meant for ambience in the recording.
Next time you’re listening, take your fore finger and ever so slightly tilt your outer ear away from your head where it meets your skull……it will come as quite a shock how massive and immediate the changes to the aural experience. This has been my argument for decades when we talk about measurements and what we know and what we don’t…..audiophilia becomes more relevant and the science of speaker design less so as there are two huge variables in the listening equation…..the 3d space and the biology of the listener…..notice how I didn’t use capability as to exclude those without the ‘skill’…..that’s where audiophoolery and the music industry went horribly wrong by not understanding their market.
Last edited:
You may be able to get the phase right in a single place in space, but in general phase will always be a mess.
dave
Audibility of phase discussion aside, there are only two polarities how you can connect it. One will have dip around crossover frequency with severely messed up phase, the other will not.
Weather you hear the difference or care is totally up to you.
https://www.diyaudio.com/community/threads/rew-phase.398657/post-7334946
My experience is that a super tweeter or any tweeter crossed over relatively high is pretty much position independent. I think the reason may be that there are so many reflections from the room in the high frequency range that the original signal gets lost.
If you feel you have problems with tweeter on axis try it on top or the sides pointing into other direction. Bettering indirect sound field.
Can do it, too.
Watch for example Pfleid FRS20 loudspeaker.
https://www.diyaudio.com/community/threads/directivity-of-fr-drivers.373003/post-6737349
https://www.diyaudio.com/community/threads/understand-clone-larsen-8.403364/post-7455502
Can do it, too.
Watch for example Pfleid FRS20 loudspeaker.
https://www.diyaudio.com/community/threads/directivity-of-fr-drivers.373003/post-6737349
https://www.diyaudio.com/community/threads/understand-clone-larsen-8.403364/post-7455502