Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Dan Haddad]

I am trying to fix the phase and alignment problems with my setup.

On tops i have JBL SRX725's, for subs i have some JR-36's
(CV jr earthquakes 18")

We have some medium size parties where we dont need the dual 18" cabs what is going to be the best way to get this setup sounding right?

i have  DRPA for setup.

[Tom Reid]

Run the Vegas from 35hz to 80hz Run the 725 from 90hz up.
Search for time alignment techniques and hope the DRPA has got enough delay to move the 725s back to align with the Vega folded horns.

[Aaron Weidner]

Run the Vegas from 35hz to 80hz Run the 725 from 90hz up.

Search for time alignment techniques and hope the DRPA has got enough delay to move the 725s back to align with the Vega folded horns.

Somewhat new around here, first time i saw somebody advise a 10hz gap in the crossover. Why? Just trying to learn. :-)

[Jay Barracato]

The electrical crossover ( numbers plugged into the DSP) is different from the acoustic crossover because the subs are run at a higher level.

Most generic crossover examples have both passbands at the same level. In practice if the kevels are not the same a small gap helps smooth the transition.

Sent from my DROID RAZR HD using Tapatalk

[Tim McCulloch]

Run the Vegas from 35hz to 80hz Run the 725 from 90hz up.

Search for time alignment techniques and hope the DRPA has got enough delay to move the 725s back to align with the Vega folded horns.

Is the the Ghost of Tom Reid Past?  IT guy and former nightclub developer?

How the heck have you been?

[Tim McCulloch]

Somewhat new around here, first time i saw somebody advise a 10hz gap in the crossover. Why? Just trying to learn. :-)

Hi Aaron-

Because we're more interested in the *acoustic crossover* than we are concerned that the electronic filters look pretty in the picture.

The acoustic crossover is the frequency range over which adjacent pass band transducers contribute roughly equal output.  We manipulate this with electronic filters and delay (because those are the tools available).  Sometimes that means overlapping or under-lapping the electrical crossover points.  The idea is that we keep BOTH acoustic contributions consistent with each other in TIME, across as much of the acoustic crossover's frequency range as possible.

[Ivan Beaver]

Somewhat new around here, first time i saw somebody advise a 10hz gap in the crossover. Why? Just trying to learn. :-)

Just to add to what others have said-but there are 2 main reasons for having a gap (but not in every case).

1: As already said-the subs are typically run hotter, meaning that the point at which the subs and tops meet would be the acoustical crossover.

This is assuming the response of the subs is flat-which is rarely the case up where the crossover point is.

2: In many cases the subs have a rising response towards the ACOUSTICAL crossover region.  So you need to cross them over lower to have a flatter response around crossover.

NOW comes the hard part.  Getting an actual proper alignment.  Without proper measurement gear and knowing how to use it to look at the phase-it becomes a real guessing game.

You can try putting in a sine wave at the crossover freq (lets just say 90Hz for now) and flipping the polarity of the tops and adjusting the delay for a null-then flip the polarity back for proper operation

HOWEVER this method has several problems.  It could be that the null method results in an improper time-that while true it give the proper time for the crossover freq-there could be notches on either side (above and below in freq) that now have notches that do not show up in the test.

The PROPER way is look at the phase response and adjust that-while also looking at the freq response to see if you are getting good addition and not putting any nulls in the response (up to an octave above and below crossover).

Or you would put some music in and simply adjust the delay on the tops until it sounds "right" to you.

Now write down the delay used and choose a different track by a different artist and adjust again.  See if you come up with the same time.

This method is best if you have 2 people-one to adjust (and not tell the person what the time is while adjusting-to keep bias out of it) and one to listen.

Be sure to try several tracks and see if you come up with the same result each time.  If not-then you need another method.

Some people are very good at this-most are not.  Just like tuning a car by listening to the engine.  In most cases it takes the proper gear to do it right.  You might be able to get "kinda close or close enough", but getting it "right" requires a bit more investment.

[Art Welter]

I am trying to fix the phase and alignment problems with my setup.
i have  DRPA for setup.

Tom's response is about the alignment I recall, and the DRPA does have enough delay, took about 6.8 ms delay on the top cabs IIRC.

[Frank Koenig]

For developing sub crossover settings, or any crossovers for that matter, I like to work from the comfort of my desk rather than the woop-woop-woop or chchchchchch of the measurement system.

I start by getting accurate measurements of the tops and subs within a couple octaves of the crossover region. For sub crossovers this is simplified by most systems (at least the dinky stuff I work with) being more-or-less omnidirectional at those frequencies, so single point (usually ground plane) measurements suffice.

For these measurements all downstream processing other than the crossover under consideration should be in place, or noted so that it can be incorporated in the subsequent simulation. Significantly, this includes the sub's excursion-protection high-pass, which typically affects the phase in the crossover region.

I then simulate the effect of various filters on the measured responses and look at both the individual phase and the summed magnitude in the crossover region. Butterworth filters of various orders are a good starting point. For a given order they differ very little from Linkwitz-Riley filters, except near the cutoff frequency, and these differences are small compared with the wild behavior of real speakers. Furthermore, Butterworth filters can have odd orders, which may give the required phase shift in the asymptotic part of the response.

The goal is to get the phases to match over as wide a range as possible surrounding the crossover and to have the summed magnitude relatively flat. Additionally, the out-of-band attenuation needs to be sufficient for driver protection (top) and avoiding response anomalies (sub). The relative delay needs to be adjusted after each change of filter settings, and I have my simulation set up to automatically calculate it. I take care not to get off by an entire cycle in the delay setting and, while this usually is obvious from the relative phase slopes, a look at the impulse responses can serve as a sanity check.

I currently believe that it makes sense to tune the subs more-or-less flat, and use additional shelving filters (one for each pass-band) upstream of the crossover if a hay stack is desired. This allows the amount (and slope) of hay stacking to be varied subjectively without messing up the sub alignment.

I also believe that separating the high and low cutoff frequencies as much as possible without getting too much of a hole in the middle gives subjectively the least muddy sound. It has been suggested to me that this is especially important when the distance between the tops and subs is large, as with flown L/R tops and centered subs on the ground.

Once the desk work is done, all that is required in the field is a quick setting of the delay using Smaart or equivalent. It is useful to note the required delay for proper alignment when the subs and tops are co-located (I call this the reference delay) so that a good guess can be made of the required delay by examining the relative distance to the tops and subs. This helps prevent mistakes in the heat of battle and may even be sufficient if there isn't time for measurement.

Time alignment is a rough business in any case as a 7 ft. difference in relative distance corresponds to a 180 deg phase shift at 80 Hz, throwing all alignment efforts out the window. I welcome your comments.
   

[Aaron Weidner]

so. much. information. I think i understand most of what is being said. I now realize why so many mobile systems sound like crap. I try my best tho. I seem to be the only one in the area that tries to pay attention when it comes to sound quality. I feel as tho this might be off topic, but, what gear is used to measure what is stated above (specifics if possible)? I've always been interested in the equipment to measure and adjust a sound system for various things but i wouldn't know where to begin.

[Caleb Dueck]

so. much. information. I think i understand most of what is being said. I now realize why so many mobile systems sound like crap. I try my best tho. I seem to be the only one in the area that tries to pay attention when it comes to sound quality. I feel as tho this might be off topic, but, what gear is used to measure what is stated above (specifics if possible)? I've always been interested in the equipment to measure and adjust a sound system for various things but i wouldn't know where to begin.

SmaartLive, USB preamp (we use Roland OctaCapture at work), Audix TM1, and the TRAINING by Rational Acoustics.  This, or Systune, work well for tuning systems in the field.  There are slightly better programs/mics/etc for lab work, like manufacturers creating files.

[Ivan Beaver]

so. much. information. I think i understand most of what is being said. I now realize why so many mobile systems sound like crap. I try my best tho. I seem to be the only one in the area that tries to pay attention when it comes to sound quality. I feel as tho this might be off topic, but, what gear is used to measure what is stated above (specifics if possible)? I've always been interested in the equipment to measure and adjust a sound system for various things but i wouldn't know where to begin.

One of the hardest things about system alignments is knowing what you can and cannot (or should not) try to "fix".

Realize that you cannot make it perfect, but trying to make it as even as you can for as many people as you can.

[Merlijn van Veen]

Without access to an analyzer I recommend using a warble (Bink test CD) instead of a sinus when attempting to align by ear.

This way you at least hear all frequencies of interest throughout the XO region.

Like Ivan pointed out, using a sinus you might very well end up being in phase for that specific frequency and still be anything from 1/2 cycle (polarity reversed) to numerous cycles off.

The result will be ripple to the left and right of the XO point. The frequency range where this ripple occurs is determined by relative level and the severity by the amount of misalignment or difference in phase slopes.

[ProSoundWeb Community]