GDan in #74 said "Also, since you are using DSP, it is very easy to adjust the frequency response of each driver to be flat at least one octave above and below the crossover frequency."
Good luck doing that to the mids on a Unity or Synergy speaker
Ι'm not sure what you mean by that.
That is not possible to equalize these speakers ?
Huh? DSPs are a god-send (in theory) for people with serious timing issues like Klipschorns with an acoustic path 20 feet behind the mids. And anybody with a separate sub cab (hint; it hardly makes much improvement, but the theory must be served).
Trivial phase adjustments are available with a DSP; but if you can hardly hear a difference with a 20 foot bass latency, hardly worth losing sleep about XO phase.
B.
How can you quote me, and give a comment completely off the topic that went with the quote?
Here's a snip of that quote and the exchange that prompted it, from earlier in this thread...
How about stopping and answering that phase rotation stuff?....
....Before i ask you about today's comment...which I can't follow either...
I'm saying "I can't follow" to be polite and see if i'm missing something necessary for understanding your posts...
It's either ask for some explanations, ...or give up and declare them total nonsense....
Interesting this came up Rob !
And yes, I agree with the difficuly of EQing a synergy/unity an octave beyond xover. Not gonna happen unless you totally manhandle the acoustic notch above high-pass.
I say interesting this came up, because I just finished another low weight synergy design, a 75x50 using a pair of 8"s from 100Hz to CD, where the acoustic notch has come in at 480Hz.
It has very nice response to 700Hz if it weren't for the notch, so right now I'm A/Bing: xover at 480Hz using the notch to assist xover,
vs crossing at 650Hz and manhandling the notch in-band.
Surprisingly, Eq'ing the notch in-band is sounding great .. both ways are sounding super...gotta keep listening to a variety be sure though.
You've lost me here, 2 channel measurements lock the time of flight of drivers, single channel and acoustic timing reference can achieve the same thing. They both need the same amount of polar measurements per driver to be able to simulate the crossover and on/off axis response.
Where can I retrieve some information about the hardening effect of LR4?It's an interresting solution, many ditch the LR4 because of the hardening effect and go for the LR2 with an pair invert.
Maybe that angle hit the listener more soft.
It sounds that you are in to this design atm. QLN reference
Last edited:
Here
Power response comparison with Linkwitz-Riley crossovers
And the bottom half of this page
Crossovers: Active crossovers
And a listen test
Hi Mark,
I've got to admit my meh project has really pushed me to up my game on measurements / crossover setup stuff.
I've been using a target curve in REW to EQ/XO to rather than trying to EQ flat and then apply the 'textbook' xo. It's working ok for me so far, slowly tweaking to get them sounding balanced in room. (I know they can sound better than they do at the moment)
I think the main hurdle is handling the horns transition to omnidirectional and making it sound natural in room.
Cheers,
Rob.
Hi Rob,
Yeah, they push me too. Mainly because the flatten out-of-band method is more problematic with them.
But i think it should be problematic, because that method only works for true minimum phase driver rolloffs.
And minimum phase can only come from a single driver.
The strong notch reflection, which is delayed in time, becomes a 'second driver' i think.
Maybe it can be corrected with FIR, which can handle time domain issues...dunno, just beginning to try there really.
So i think your approach to matching to a curve is on the money.
And yes, totally agree about balancing the transition of pattern in a room.
I've got a 60x60, a 60x40, a 90x60, and a 75x50...most of them with removable secondary flares.... to testify to the difficulty of that hurdle.
Haha...if they didn't sound so damn good, I'd say screw it.
But there's just something about closely located acoustic centers...or so i think i hear, and believe.
Yeah, they push me too. Mainly because the flatten out-of-band method is more problematic with them.
But i think it should be problematic, because that method only works for true minimum phase driver rolloffs.
And minimum phase can only come from a single driver.
The strong notch reflection, which is delayed in time, becomes a 'second driver' i think.
Maybe it can be corrected with FIR, which can handle time domain issues...dunno, just beginning to try there really.
So i think your approach to matching to a curve is on the money.
And yes, totally agree about balancing the transition of pattern in a room.
I've got a 60x60, a 60x40, a 90x60, and a 75x50...most of them with removable secondary flares.... to testify to the difficulty of that hurdle.
Haha...if they didn't sound so damn good, I'd say screw it.
But there's just something about closely located acoustic centers...or so i think i hear, and believe.
Depends on what you mean by mitigate...
if you mean at a single freq, yes you can. (But choose the single freq at the expense of all others.)
if you mean through the entire xover range, no you absolutely cannot.
Hi manninen,
REW doesn't need a umik, works fine with any old mic.
I have a umik but never use it, preferring an analog mic with soundcard loopback for timing reference.
A single REW loopback measurement, when there is sufficient hf content, consistently gives the same timings as dual channel Smaart, ime.
REW's acoustic timing reference works well too, but not quite as well as loopback, especially when testing lower freq drivers without sufficient higher freq content.
By sufficient hf content, I'd guesstimate 1000Hz for REW loopback, and maybe 2-3000Hz for acoustic timing. Pure guesses though...i don't use REW for timing as single channel is just too much of a pain to use for dialing in phase, imo.
But must say, it is pretty dang time accurate ...
Last edited:
REW is such a versatile tool. I use(d) it to EQ my speakers, but could never get the timing to work reliably. Furthermore, my living room has the accoustics of an empty Olympic swimming pool so rather less incentive to measure time. There are work-arounds that will dial in timing, maybe not perfect, but still producing nice results. It's not entirely clear that a 1 KHz burst of ten mSec, looped, is the best way to set timing, but I can tune it by ear.
Mark100 is right about the Unity/Synergy "notch." Per the original design, the mids are supposed to be crossed somewhat under the notch, which is typically 2x the crossover frequency of the mids to the highs. And, at least based on commercial Unity designs (e.g. Yorkville), the mids may be closed-back, probably to maximize efficiency, but also this limits their low end substantially. I'd estimate my Yorkville U15 mids drop quickly below 300 Hz, which is their x-over. It's my understanding that Mark100's or similar "home" horns can use mids or even woofers for the "mids" and I guess, trade maximum SPL for frequency response.
Mark100 is right about the Unity/Synergy "notch." Per the original design, the mids are supposed to be crossed somewhat under the notch, which is typically 2x the crossover frequency of the mids to the highs. And, at least based on commercial Unity designs (e.g. Yorkville), the mids may be closed-back, probably to maximize efficiency, but also this limits their low end substantially. I'd estimate my Yorkville U15 mids drop quickly below 300 Hz, which is their x-over. It's my understanding that Mark100's or similar "home" horns can use mids or even woofers for the "mids" and I guess, trade maximum SPL for frequency response.
Last edited:
I don't have a UMIK either, it's the difference between running a loopback from the soundcard to get the timing reference or an acoustic signal to do the same thing. Both work with an analogue mic and soundcard but only the acoustic one works with USB mics.
Try it for yourself and find out, I have nothing setup to measure at the moment.
Does the phase not correspond to the magnitude?
What made you think that?
BTW I think talk of synergies has nothing to do with this thread.
Here
Power response comparison with Linkwitz-Riley crossovers
And the bottom half of this page
Crossovers: Active crossovers
And a listen test
The first link is showing the difference between LR2 and LR4 when there is a directivity mismatch between drivers. If you don't want a waveguide on your tweeter to help match the directivity then it's a reasonable comparison but mainly highlights how hard it is to get a tweeter and larger mid driver to work together.
In the second link the "hardening" is talked about in relation to listening to music with the lower or higher order filter in place with one driver.
It's not too hard once you know the difference between room/diffraction effects and speaker response. Can get closer to linear phase if you set your delays right. This works best with a coaxial, mid/tweet fullrange, or low crossover tweeter.
Flatten your drivers passed their intended passband, delay your mid relative to your woofer a cycle or two, do the same for the tweeter. This was done in about 10 minutes. It can be optimized surprisingly well without FIR.
Flatten your drivers passed their intended passband, delay your mid relative to your woofer a cycle or two, do the same for the tweeter. This was done in about 10 minutes. It can be optimized surprisingly well without FIR.
Attachments
Last edited:
Thread starter has an umik, how do you loopback timing with that?
Sure i know how acoustic timing works
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.