Going Tri-amped...What About Delays?

[Zooberwerx]

Well, once again, I have too much time on my hands and a 100' roll of 13 g 4-conductor cable. Thought I would try my hand at tri-amping two biampable 3-way mains and 18" subs. I'll be using a dbx DriveRack's xovers set at 100 hz and 630-750 hz (thank you greenboy). The DR also has delay capabilities for each of the bands (high, mids, lows) which can be adjusted in terms of time (milliseconds) or distance (feet) to compensate for sonic anomalies. As you can guess, I don't have a strong grasp of the concept and in search of a starting point.

FWIW, the dbx delay structure smacks strongly of the BBE Sonic Maximizer claim whereas fixed xover points / fixed delays are introduced for correction. The dbx parameters are, of course, user-adjustable.

Riis

 

[D.A.R.K.]

what are your top boxes?
the mid/ high crossover point sounds a bit low imho,
but it really depends on the drivers/ design. the slope comes into play with that as well, of course.
the delays in the drive rack are indeed quite useful for time alignment, but i would consult the manufacturer,
ideally the cabs were designed to take alignment into account
(ideally, heh) but you could experiment with some mathematics and the fine adjustment setting.
one thing you could try for fun would be delaying the entire input to
line up the rig with your back line, just measure the distance
from the cabs and do the calculations.
ime delay to back line not only lines up the sound waves from
stage amps with foh, but can also help lessen feedback issues
with foh by putting time in between the source and system reaction to that source. we are talking milliseconds, i use these standard techniques but the rigs and rooms are typically larger... still, same physics...

 

[Zooberwerx]

what are your top boxes?
the mid/ high crossover point sounds a bit low imho,
but it really depends on the drivers/ design. the slope comes into play with that as well, of course.
the delays in the drive rack are indeed quite useful for time alignment, but i would consult the manufacturer,
ideally the cabs were designed to take alignment into account
(ideally, heh) but you could experiment with some mathematics and the fine adjustment setting.
one thing you could try for fun would be delaying the entire input to
line up the rig with your back line, just measure the distance
from the cabs and do the calculations.
ime delay to back line not only lines up the sound waves from
stage amps with foh, but can also help lessen feedback issues
with foh by putting time in between the source and system reaction to that source. we are talking milliseconds, i use these standard techniques but the rigs and rooms are typically larger... still, same physics...

The cabs are Carvin LS1503's (15", 5" midrange, HF horn). The subs are budget Yorkville 18"s.

I'm having some difficulty understanding exactly what we're aligning...the response amongst the different PA drivers, PA and backline amplification, or a combination of both.

BTW, if you can suggest a more appropriate xover point between the mids and highs (630-750), I'd greatly appreciate it.

Thanks

Riis

 

[craig.p]

You are aligning the delivery times among all four PA drivers, so that a sub signal and a 15" signal and a 5" signal and an HF signal all will arrive at the listener at exactly the same time. Conceptually, consider a widely-spaced chord played on an organ, with organ-pedal bass, too. You'll have energy coming out of all four drivers when that chord is played. You want it all to arrive at the listener's ears at the same time. We're talking only tiny fractions of a second, though, and there's always been vigorous debate whether worrying about time alignment is even worth it in a live (and often noisy) setting. Consider that, along with the type/quality/distortion content of the music that's going to be played through that kit, the alcoholic content of the patrons' bloodstreams, and the general appreciation level of the audience for audio clarity. I'm not saying don't go though the exercise, just don't overvalue its return in certain circumstances. There may be more net audible benefit to spending your time elsewhere in the signal chain.

 

[craig.p]

I wanted to add something I forgot, and I apologize for forgetting. There's also the concern about driver overlap in the crossover region. If two drivers are contributing to a frequency at or perilously close to a crossover point, you want them in phase, i.e. non-canceling. (The crossover order will make a difference here, but that probably goes a little off topic so I'll mention it only in passing.) Again, will phase-induced suckout at crossover frequencies be audible by your particular audience? Maybe, maybe not. As I said before, there's lots of other stuff going on in a live setting that can benignly mask it.

 

[Zooberwerx]

There may be more net audible benefit to spending your time elsewhere in the signal chain.

Thanks for the insight, I have to agree. We've been able to pull of more than a handful of successful bi-amped shows with a simple dbx crossover...no adjustable delays or slopes. All I really know or understand about delays is that (a) they're available on the DriveRack and (b) in the right hands, they may improve the overall sound quality.

Riis

 

[GregShadoan]

Time alignment is important. Some boxes sound horrible without it. Most big box houses have their own built in processing now, to take this all into account.
VRX, self powered with DBX and crown internal etc.. for example. VDOSC has its own set of rules, and such. The carvin / yorkville combo probably could sound just fine without it. To use the alignment tools, you will need to measure the times, and the separate bands, and adjust until they are all the same, is the theory. Most sound people with a smarrt system could do it for you. The good news is you really only have to do it once. Given that you setup the system consistently and have the timing programmed into the crossover, you should be good to go.

 

[WayneP]

I'm having some difficulty understanding exactly what we're aligning...the response amongst the different PA drivers, PA and backline amplification, or a combination of both.

To elaborate a bit on what craig.p said about the delivery times of the drivers, just look at a horn in a cabinet vs. a cone speaker. The horn is deep, which puts the compression driver a few inches behind the voice coil of the cone driver. So, the sound from the latter will reach your ears a bit sooner than the sound from the former. That’s what the time alignment is trying to correct.

Regards,
Wayne A. Pflughaupt




Pedulla Club #45

 

[craig.p]

Some history: I did an experiment when I first started biamping. This was in approximately 1975, and it was with (hold your breath...) Altec 1204Bs. Can't recall the crossover I was using, but it was a 1U two-way, second-order (12 dB/octave) and it had a continuous-delay feature and a polarity feature. The poor man's way of setting it was to inject a very low-volume sine wave into the mains buss at the crossover frequency, and then to adjust the crossover's polarity switch and delay knob so the tone seemed to be coming from a point equidistant from the LF driver and the horn (rather than from only one or the other, or from both but with silence in between them). Once you accomplished this, it meant that both LF and horn were phased so that the energy they contributed at that frequency was additive. It was crude and imperfect, but it worked reasonably well -- at least "getting the grossness" out of driver phasing.

My experiment? I did it that way one night, and then the following night (same room, same 250-300 customer capacity, both nights packed) without telling a soul I purposely dialed the delay to be completely wrong so there was a giant 500 Hz null. Guess what. NO ONE KNEW THE DIFFERENCE. Not even the band's manager, and he was always paying attention to our live sound.

So, am I saying phase isn't EVER important? Not at all. What I am saying is that you need to determine when it IS important to nail it and when there'll simply be no payback on the effort. Every room's different. And again, by "room" I mean room itself, clientele, average per-carcass BAC , music type, etc., etc.

EDIT: If you're always using the same speakers, then once the phasing exercise is done, it's done for good, just as Greg says, as long as the relative physical positions of the SUB/LF/MF/HF drivers never change. (Just thought I'd better add that.)

 

[Bob Lee (QSC)]

If the drivers are not reasonably well aligned in a vertical plane (or the plane perpendicular to the intended direction of the sound), then you'll need to time align the frequency bands so you don't get response aberrations and lobing at the crossover points.

It's more important at higher-frequency crossover points; for example, a 500 µs misalignment (approx. 6.6 inches) at 100 Hz is an 18° offset—not optimal but not terrible, either—but it's 180° at 1 kHz.

 

[TimmyP]

To add to what Bob said, if it's not right, the box may sound fine on axis, and terrible off axis. It's not a simple process.

 

[Zooberwerx]

Okay, let's put the delay discussion on the back burner for a moment...

I'd mentioned earlier that, per suggestion, the mains would be bi-amped @ 630-750 hz (the box's passive xover freq's are 500 hz and 3.3 khz). I'm now hearing this may be a bit low given the intended application. What xover point(s) would be appropriate and why? Reminder: this is a tri-amped system....the subs and mains are xover'd @ 100 hz.

Thanks again!

Riis

 

[GregShadoan]

Well, without any scientific measurement equipment, or a proper chamber, you are just going to have to listen to them. An experiment you may want to try would be to A/B them. Set one box in a passive configuration, and the other in a bi-amp configuration. Sweep the crossover point and phase around on the 15"/6"-1" point and see what sounds best.
Here's the thing. Even if you tri-amp the rig, the high element is still passive, and mismatched ie. a 6" and a 1" with presumably no way to align them other than the passive crossover. So it will still be a crap shoot. Actually, to do this right, you would have to make this a 4 way rig, and crossover/align/phase/eq/power each of the 4 bands separately. Other than doing that you can get it close, and align/phase/eq the "3" bands that you will have control over and call it a day. You will get some benefit by tri-amping this system, mainly headroom. You should be able to get a 20-30% by removing the passive crossovers in system. You are still going to be using the 6"-1" passive crossover 3.3K point. It is what it is.
I would recommend the A/B method (taking the subs out of the equation) with some well produced CD's and just adjust the filters to taste. When you feel you have optimized the mid/high, then introduce the subs.
Good luck, I wish I could come and help out. Sounds like a fun experiment.

 

[Zooberwerx]

I would recommend the A/B method (taking the subs out of the equation) with some well produced CD's and just adjust the filters to taste. When you feel you have optimized the mid/high, then introduce the subs.
Good luck, I wish I could come and help out. Sounds like a fun experiment.

This I can do. I'll have to be super-careful with the HF as I'll be assigning am RMX 1450 which pushes 280 watts / 8 ohm. I'll use an RMX 2450 on the MF / 15"s.

Riis

 

[Bob Lee (QSC)]

Want to know what's even more telling? Pink noise. It's just not as enjoyable to listen to.

But you can hear changes in phase, timing, lobing, reflections, etc., very clearly.

 

[Zooberwerx]

Want to know what's even more telling? Pink noise. It's just not as enjoyable to listen to.

But you can hear changes in phase, timing, lobing, reflections, etc., very clearly.

I can do that, too! I use it for setting gain, headroom, limiters, etc.

Riis

 

[Passinwind]

Okay, let's put the delay discussion on the back burner for a moment...

I'd mentioned earlier that, per suggestion, the mains would be bi-amped @ 630-750 hz (the box's passive xover freq's are 500 hz and 3.3 khz). I'm now hearing this may be a bit low given the intended application. What xover point(s) would be appropriate and why? Reminder: this is a tri-amped system....the subs and mains are xover'd @ 100 hz.

Riis, do your cabs have the stock Carvin drivers in them?

BTW, when I reverse-engineered greenboy's 1503 crossover I found it to be a bit different than the Carvin specs might lead you to expect. The details are on my old computer though, and I can't recall the particulars off the top of my head.

 

[Zooberwerx]

Riis, do your cabs have the stock Carvin drivers in them?

BTW, when I reverse-engineered greenboy's 1503 crossover I found it to be a bit different than the Carvin specs might lead you to expect. The details are on my old computer though, and I can't recall the particulars off the top of my head.

Yes, the cabs have the original drivers. Sounds like you have an idea brewing.

Riis

 

[walterw]

hey riis, you're plunging headlong into the woods with bandpass timing delays, but do you even have a master EQ yet (besides the driverack?)

 

[Zooberwerx]

hey riis, you're plunging headlong into the woods with bandpass timing delays, but do you even have a master EQ yet (besides the driverack?)

Headlong into the woods?....try "off a cliff". Yes, I have an older ART dual 15 GEQ (free!) that I use as for rescue. The DR's auto-EQ is usually more than adequate and results are reproducible venue-to-venue.

See you later this AM...I have a power amp goin' south. Need to see your guys across the aisle.

Riis