Print

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

[Michael Lawrence]

It's such a basic function, any DSP that can't get even this right isn't something I'd use.

You may find it surprising that there is no polarity control anywhere in the d&b audiotechnik ecosystem despite the fact that they are one of the market leaders in loudspeaker systems. 
I spec a GALAXY in front of everything so I am not stuck without any of the basics. (Some other leading manufacturers omit high pass / low pass / shelving filters)

[Jose Peralta]

I know the dbx VENUE 360 has all these feature in their interface control, polarity inversion, phase, and delay.

[Mike Caldwell]

If the system is properly time/phase aligned, a polarity reverse should result in a deep notch at the acoustic crossover when the mains and tops are equidistant and tested near field.

Also can be used as trick for finding driver delay settings.

[John Schalk]

I know the dbx VENUE 360 has all these feature in their interface control, polarity inversion, phase, and delay.

Yes, and I think it's the only single space driverack unit that has polarity and enough PEQ filters to allow you to really tune a speaker system.  I tried using a PA+ unit last year to implement EAW's recommended DSP settings for bi-amping the JF260z, which requires a polarity change on one output, and found out about the limitations of that MI level piece of gear.  It also doesn't have enough PEQ horsepower to handle EAW's not very complex suggested DSP settings for the 260z.  The main and matrix outputs of an X/M 32 have more DSP processing capability than a Driverack PA.

[Tim McCulloch]

You may find it surprising that there is no polarity control anywhere in the d&b audiotechnik ecosystem despite the fact that they are one of the market leaders in loudspeaker systems. 
I spec a GALAXY in front of everything so I am not stuck without any of the basics. (Some other leading manufacturers omit high pass / low pass / shelving filters)

This.  l'Acoustics, I'm calling your name here.

I'm now SE for a K3 installation.  I'd take some common system tools like the filters Michael mentions over the "zoom" feature.  I don't care about making the K3 array sound like vDosc at 30 meteres or other silliness.  I want the rig to sound good in the room, not sound like a different rig in another space.

As I'm new-ish to this product, I presume things I don't like are a result of my errors or lack of familiarity, but that only goes so far when the tools and techniques I use are not available to me.  I really, really wish this client had consulted me before making the commitment to the design they purchased.

[Jose Peralta]

Yes, and I think it's the only single space driverack unit that has polarity and enough PEQ filters to allow you to really tune a speaker system. 

John, thank you for the response. That is what I am looking for in a loudspeaker management system that you are able to control for convenience.

[Peter Kowalczyk]

As someone noted, it's common for the subs to naturally lag behind the tops, and so it's more likely that the tops need will a little delay to line up with the subs.  But, it's very difficult to start your time alignment with a Sub measurement, since the impulse response is so long.

So, my approach has been to
- start by adding, say, 10 ms of delay to ALL processor outputs equally.  This gives me a buffer to enable 'negative' relative delay later.
- measure the tops, and set the delay finder accordingly. 
- Mute tops, and measure subs.
- Now, you can REDUCE the delay on the subs (and flip polarity if necessary) to get the phases to play nice*.
- Once you're happy with the time alignment of the whole system (including your fills or whatever other zones you have), subtract the smallest delay value from ALL output channels, so the relative delay between channels remains, but that pre-emptive bulk delay 'buffer' is removed.

* Getting the phases to play nice often takes more than just delay.  Typically, one polarity setting is a clear winner.  An out-of-band low-pass filter (e.g. in the 160-400 Hz range) can modify the subwoofer's phase slope in the actual crossover region.  Naturally, the high pass behavior of the Tops (whether baked into factory processing, or added by you) does the same on it's side of the crossover.

If you're measuring JUST subs, (e.g. to align a cardioid array), then I'd suggest to manually enter and adjust the measurement delay in Smaart until you get a maximally-flat phase trace through the sub's operating band.  It should look like a 'smile.'  This won't help your sub-to-top phase alignment, but it makes sub-to-sub comparisons easier.

Note that differences in speaker processing systems (analog vs. older DSP vs. newer DSP) can result in significant differences in processing latency.  If your system contains speakers with different native processing systems, then aligning those speakers could require 'non-intuitive' processor delays to compensate.

[Jose Peralta]

As someone noted, it's common for the subs to naturally lag behind the tops, and so it's more likely that the tops need will a little delay to line up with the subs.  But, it's very difficult to start your time alignment with a Sub measurement, since the impulse response is so long.

So, my approach has been to
- start by adding, say, 10 ms of delay to ALL processor outputs equally.  This gives me a buffer to enable 'negative' relative delay later.
- measure the tops, and set the delay finder accordingly. 
- Mute tops, and measure subs.
- Now, you can REDUCE the delay on the subs (and flip polarity if necessary) to get the phases to play nice*.
- Once you're happy with the time alignment of the whole system (including your fills or whatever other zones you have), subtract the smallest delay value from ALL output channels, so the relative delay between channels remains, but that pre-emptive bulk delay 'buffer' is removed.

* Getting the phases to play nice often takes more than just delay.  Typically, one polarity setting is a clear winner.  An out-of-band low-pass filter (e.g. in the 160-400 Hz range) can modify the subwoofer's phase slope in the actual crossover region.  Naturally, the high pass behavior of the Tops (whether baked into factory processing, or added by you) does the same on it's side of the crossover.

If you're measuring JUST subs, (e.g. to align a cardioid array), then I'd suggest to manually enter and adjust the measurement delay in Smaart until you get a maximally-flat phase trace through the sub's operating band.  It should look like a 'smile.'  This won't help your sub-to-top phase alignment, but it makes sub-to-sub comparisons easier.

Note that differences in speaker processing systems (analog vs. older DSP vs. newer DSP) can result in significant differences in processing latency.  If your system contains speakers with different native processing systems, then aligning those speakers could require 'non-intuitive' processor delays to compensate.

Very good information. Yes, I do set tops first when I delay locate then the subs. One of the questions I want to answer is where would be the appropriate location for the measurement mic to be place when you are doing a single measure at a time. I would assume foh center? If I have single mic for this, where would it be ideal and when I do sub measurements, how can I get an impulse response to see where the peak is for sub bass frequencies?

[Frank Koenig]

Very good information. Yes, I do set tops first when I delay locate then the subs. One of the questions I want to answer is where would be the appropriate location for the measurement mic to be place when you are doing a single measure at a time. I would assume foh center? If I have single mic for this, where would it be ideal and when I do sub measurements, how can I get an impulse response to see where the peak is for sub bass frequencies?

First, forget about the impulse response. It’s a great representation for archiving measurements but of little use for aligning subs. What you’re trying to do, at the acoustic crossover frequency as determined by measurement, is to match the phase and the slope of the phase. The former is necessary to prevent cancellation and the latter to match the group delay. The group delay is the delay of the envelope of a time-limited, narrow band signal. Think of 7 cycles of 80 Hz surrounded by silence, for example.

The tools you have to do this are the order (1st, 2nd, 3rd, etc.) and type (Butterworth, Linkwitz-Riley) of filter, the cutoff (-3 dB corner) frequencies, relative polarity, and delay. Higher order filters (high-pass or low-pass) produce more phase lag with increasing frequency which tilts the phase trace down more through the transition. Different types of filters of the same order have slower or faster transitions but are identical in the pass- and stop-bands. Polarity gives you 180 deg of phase shift with no time (phase slope) change, and delay tilts the entire phase trace down. A lower crossover frequency increases the size of the sweet spot in proportion to the increase in wavelength.

It may be useful to bias the region of phase overlap to the high side of the crossover frequency. This way when the relative level of the subs is increased for subjective effect, shifting the crossover frequency higher, alignment is better maintained.

As for mic position, on the ground near the center of the primary listening area is good. If you must put the mic up in the air be aware of the dip caused by the ground bounce cancellation. This can be verified by observing that the frequency of the dip changes with the height of the mic and that the frequency is close to that estimated from the difference between the direct path and the bounce path.

In a typical small system with left and right mains I mute one side during the alignment to prevent interference between the mains from confounding the measurement. Take into account the increase in mains level when both sides are unmuted.

When all is said and done and the bass still sounds muddy it can sometimes be salvaged by opening up the gap between the low- and high-pass. I’ll take a little magnitude dip over mud.

There is more to this as others will surely point out. This is just a little nutshell.

--Frank

[Jose Peralta]

What you’re trying to do, at the acoustic crossover frequency as determined by measurement, is to match the phase and the slope of the phase.

Exactly. Would it be easier to use wrap phase trace reading or delay locate so the phase trace is flat at the crossover point? I pretty sure with the latter will be easier since if I need a polarity inversion due to crossover slopes configurations, it would be easier to read.

[ProSoundWeb Community]