SRX-712M Biamp DSP tunings

[agedhorse]

As I promised a while back, I am sharing a set of parameters that I generated while developing a tuning for SRX-712M's used in Biamp mode for stage wedges, While I know this is a bit excessive of an application of this box for most of you, it's a pretty common box on the higher level systems and these tunings will work well for speaker on a stick WITHOUT SUB. If I get some time, I have an SRX-718M sitting in the shop and I'll try to run it.

 

This was developed while testing a (replacement) BBE DS-48 which I had to fix before programming. Microscopic solder bridge on the DSP causing intermittent operation. Turns out their PCB hygine is not as good as it should be. With lead free solder chemistry, cleanliness is absolutely critical as dross captured in the surface flux/oil in the solder bath (it's not purely reflowed) can easily contaminate the top surface. Grrr.

 

Lows

HPF = 49.6Hz, 24dB, BW

LPF = 1.21kHz, 24dB, LR

Gain = +3dB

Polarity = +

Delay = 0.50msec

PEQ:

315Hz, Q=2, G=-3.9dB

231Hz, Q=2, G=+3.0dB

600Hz, Q=3, G=+1.6dB

72.9Hz, Q=1, G=+5.0dB

105Hz, Q=3, G=-2.0dB

 

Highs

HPF = 1.21kHz, 24dB, LR

LPF = thru

Gain = -7dB

Polarity = +

Delay = 0

PEQ:

2.16kHz, Q=3.8, G=+2.4dB

2.88kHz, Q=3.0, G=-3.0dB

15.4kHz, HF shelf, G=+4dB

 

There are some tricks I specifically used with cascaded filters modifying each other's response at their margins (I use this technique in my analog filter design in some pretty popular products as well, so it's not something unique to DSP) to gain performance improvements not available by using a single filter section. This is generally used to overcome a mechanically induced response anomoly.

 

There is an intentional gentle rolloff from flat starting about 2kHz and ends up -2dB at about 15kHz where the HF lift filter transitions. This is also common, especially in near-field applications where you may get loud. There is compound filtering on the LF section to improve performance without a sub while still maintaining reasonable protections under high drive conditions. I would not power this box at more than 800 watts RMS, this is my disclaimer and my linearity testing confirmed this.

 

This is good palce to start, then add a little 1/3 octave to touch up for room conditions (if needed). It also reinforces my experience that this is one heck of a nice box with an outstanding HF section.

 

I think my tunings for THIS box is better than Harman's, there is something goofy with the 712 model that the 722 tuning does not share. I don't know why they underlapped and used such disparate crossover frequencies. There was NOTHING in the basic response that would even begin to indicate that it was necessary.

 

Anybody biamping this box, give it a try and report back. This is the sort of stuff I do as my day job, it's not quite as easy as it seems from the above numbers. There is quite a bit of interpretation of the data that's required to make good decisions as to what's good data, what's the best and most practical way to proceed, and when do the solutions start getting in the way of the problems.

[sibyrnes]

The JBL tunings for the 712 show a polarity reverse for the highs. Your's show = + which I assume is the opposite. Could you please explain? I'm also curious why you compare the Harmon tunings of the 712 to the tunings of the 722 since they use a totally different high freq driver and horn.

Thanks.

[jlb]

Thanks for the info Aged!

 

Would love to see what you come up with running passive with a 718.

[agedhorse]

 

The JBL tunings for the 712 show a polarity reverse for the highs. Your's show = + which I assume is the opposite. Could you please explain? I'm also curious why you compare the Harmon tunings of the 712 to the tunings of the 722 since they use a totally different high freq driver and horn. Thanks.

 

 

I found that I was able to achieve better correction with delay of the LF element only. The polarity reversal probably works (or worked) better for them with their goofy choices of severe underlapping and very shallow slope on the HF element. What it looks like to me is that somebody started down the path with goofy asymmetrical widely spaced crossover points, perhaps for an initially sound reson or possibly due to an error early on in the development process, then it got carried through to where it ended up being corrections on corrections.

 

The reason I compared is that while the drivers are different, they are in the same class (large format), and the crossover frequency/slope/level (assuming there's adequate HF capacity) is generally chosen based on low frequency driver dispersion (or beaming) characteristics. Because I got the calculated results, and the the performance is dead on with what I expected, I have good confidence that this is a good choice for this product. I could have just as easily chosen to compare with the 715, which uses the same HF driver, which also uses the same basic crossover characteristics that I calculated to be good.

 

With regards to the delay and polarity reversal, this is the basis of 2 different ways to achieve the same thing. I have no problem with reversing the polarity of the HF device and adjusting the delay, and in fact may result in virtually identical measurements. As the wavelengths shorten, absolute polarity becomes irrelevent and the importance becomes meaningful around the crossover point. In fact, for stage monitor use which is generally pretty much a near field, on axis application, I will argue that phase coherency at the crossover point is meaningfully important (in agreemenmt with Dookie here).

 

This is a good example of where years of experience doing this for manufacturers pays off. Sometimes, when you see something that doesn't make any sense, it's because it doesn't. The results also (subjectively) sound better.

[RoadRanger]

Hey, remember when we in the USA had better engineering than that? I mean WTF JBL ?

[sibyrnes]

Thanks for the reponse, agedhorse. I an very interested in what you have to say because I am a big fan of the 712. I almost always use them passive. I have used them as part of a tri-amped system using an analog JBL crossover(no delay or polarity reversal) and was not impressed with the results.

[drumstix]

Thanks for the info Aged. If at all possible, can you play with a 712 in the passive setup? (non bi-amped) I would like to see what kind of cuts and boost you would come up with. I will be running a pair of these, one per side for a speaking gig next month.

[agedhorse]

 

Thanks for the reponse, agedhorse. I an very interested in what you have to say because I am a big fan of the 712. I almost always use them passive. I have used them as part of a tri-amped system using an analog JBL crossover(no delay or polarity reversal) and was not impressed with the results.

 

 

In passive mode there's no delay of course but there may be HF polarity reversal. I haven't looked. If I have to go passive, I'll just take what they give me as it's pretty darn useable.

 

With active crossovers, the attention is in the details. Don't know what the frequency or slopes were, any eq that may have been used (or abused) and the balance between the bands. All of this is really important. I have used them triamped with 718's and they sound really nice, especially for the size. The 712's limiting factor is LF performance and below 60Hz they suffer like most compact speakers do.

[dennis a]

Were the tunings listed above developed specifically for the BBE DS-48? If so, did you try loading them into something like a DR260 to compare? Just curious.

[agedhorse]

 

Were the tunings listed above developed specifically for the BBE DS-48? If so, did you try loading them into something like a DR260 to compare? Just curious.

 

 

Funny you mention this, I just did this with triamp and a 718 sub on my standard 260 test set-up.

 

The problem with directly comparing is that the measurement environment is not identical, nor is the position of the top speakers relative to the bounderies so there is some boundery interference differences with the second set of measurements. I also ended up prefering a little less HF output, I might knock a dB or two off of the HF output on the DS48 as well.

 

Some of this comes down to personal preference, and of course moving the measurement mic even a few inches changes the results, which is why I say that spending an inordinate amount of time trying to eek the last gnat's ass out of a speaker is likely to be an excercise in futility... all you are doing is stirring the compromise and trade-off pot.

[agedhorse]

Here's what I got using a DR-260 and triamp w/ 718 on the bottom and 712M on the top. Using QSC PLX-3402 on subs and lows, PLX-1202 on highs.

 

Subs

HPF = 31.3Hz, 24dB, BW

LPF = 91Hz, 48dB, LR

Gain = +3dB

Polarity = +

Delay = 0.00msec

PEQ:

41.7Hz, Q=5.01, G=+1.5dB

66Hz, Q=7.83, G=-2dB

Easyover off, threshold=+6dB

 

 

Lows

HPF = 91.3Hz, 48dB, LR

LPF = 1.22kHz, 24dB, LR

Gain = 0dB

Polarity = +

Delay = 0.00msec

PEQ:

648Hz, Q=3.876, G=+2.5dB

1kHz, Q=1.573, G=+2.5dB

354Hz, Q=1.382, G=-2.5dB

Easyover off, threshold=+3dB

 

Highs

HPF = 1.21kHz, 24dB, LR

LPF = thru

Gain = -15dB

Polarity = -

Delay = 0

PEQ:

3kHz, Q=2.996, G=-1.5dB

2kHz, Q=5.706, G=-2.0dB

5.04kHz, Q=4.41, G=-1.5dB

11.6kHz, Slope=9, G=-5.5dB (hf lift)

Easyover off, threshold=-16dB

 

As you can see, a different set of parameters that maintain quite a bit of similarity are present, I preferred the HF horn down a few dB more on this set-up, perhaps it's a slightly different environment or just a different preference day for me. I generally do not like a flat HF response all the way out, I like a gradual rolloff of a couple of dB from about 2.5k-15k, just is a little more forgiving and my ears don't do the rolloff for me;)

 

Since this is a different environment with some significant bounderies, I had identified the bounderies and their effects the best I could so the results are reasonable and corrolate reasonable well. It's interesting that in this case I found that inverting the HF driver worked better (actually a lot better) than the delay, and there was no meaningful difference delaying the HF driver here... and I don't know why, but both methods are reasonable but like I have always said, changing the measurement environment (including mic position by a little bit) can totally change the results, so there is considerable latitude in the interpretation and application of the basic measurement/correction techniques. There are some differences in measurement mics, testset calibration, mic calibration, etc. There are also major boners out there that are just goofy and wrong when looking at the overall solution. Exact is not necessary, close is good enough in 99.9% of the cases because exact changes with many ofthe variables and if you don't have control of things like room acoustics, it doesn't matter what you would like to do, it is what it is.

 

Now that I have my universal test system set up, I'll see if I can run a passive 712 test, that would be on top of a sub since that's how most are run. That also means that there will be a significant boundery event around 150-ish Hz which I will ignore with my parameters since that will be different depending on the surface boundery material.

[agedhorse]

Here's the best I can do with a passive 712M. The out of the box flat eq has a fair amount of 11-15kHz going on, kind of the bright JBL thing which I do not care for myself. The biggest improvement here is the reduction of this. For biamped w/ subs, ignore the lowest band, that's a LF extension. I used a HPF of about 62Hz, BW-24 because these boxes don't excel at LF reproduction, but would kill for acoustic and jazz without subs if that's the style.

 

PEQ:

81Hz, Q=5.706, G=+4.5dB (note this works in conjunction with the HPF)

225Hz, Q=4.41, G=+4dB (note some of this is boundery cancellation so flavor gently)

500Hz, Q=6.491, G=-3dB

1.68kHz, Q=2.035, G=+3dB

14.7kHz, Q=5.706, G=-5.5dB (HF response taming, they use a passibe lift network that's excessive IMO)

 

Also, I used borrowed PEQ band to achieve the fifth band. Most of you using subs will not need band 1, it's mostly for informational purposes.

 

Gotta get back to work!!!

[drumstix]

Thanks Aged !!!

[jlb]

So in a biamped 712/718 setup would you still use 91Hz, LR 48, db as both your subs LP and tops HP filters? Why the steeper slope at the crossover point? You previously recommended a LR 24 octave filter.

 

Looking forward to trying these settings. Thanks for the update Aged!

[dennis a]

 

Here's what I got using a DR-260 .

 

 

Thank you

[agedhorse]

 

So in a biamped 712/718 setup would you still use 91Hz, LR 48, db as both your subs LP and tops HP filters? Why the steeper slope at the crossover point? You previously recommended a LR 24 octave filter. Looking forward to trying these settings. Thanks for the update Aged!

 

 

Either 24 or 48dB/octave is viable, JBL recommends on the PRX-618XLF and the transition within the crossover area is a bit better IMO w/ 48dB/octave but either is acceptable.

 

There are no exacts but a range of acceptable solutions. This is what I was trying to say, AND there are a range of poor solutions as well. Sometimes anechoic flat is not ideal because it does not represent how the product will end up being used 99% of the time BUT it's a good start. I went ahead and let some typical room interactions enter my solution because it's what most of you will see. I also find a pure flat response at the very high frequency band a little too much for most applications, over the years I have found most folks to prefer a gentle coupld of dB rolloff but if you have hearing challenges, this may not be either audible or desired but I am looking at the general audience type population not somebody who has hearing damage.

[drumstix]

 

Either 24 or 48dB/octave is viable, JBL recommends on the PRX-618XLF and the transition within the crossover area is a bit better IMO w/ 48dB/octave but either is acceptable. There are no exacts but a range of acceptable solutions. This is what I was trying to say, AND there are a range of poor solutions as well. Sometimes anechoic flat is not ideal because it does not represent how the product will end up being used 99% of the time BUT it's a good start. I went ahead and let some typical room interactions enter my solution because it's what most of you will see. I also find a pure flat response at the very high frequency band a little too much for most applications, over the years I have found most folks to prefer a gentle coupld of dB rolloff but if you have hearing challenges, this may not be either audible or desired but I am looking at the general audience type population not somebody who has hearing damage.

 

 

 

 

Since you brought up the PRX-XLF sub Aged, have you ever had the chance to compare it to your one of your SRX718's?

[agedhorse]

 

Since you brought up the PRX-XLF sub Aged, have you ever had the chance to compare it to your one of your SRX718's?

 

 

No, but everything I have seen makes me believe thay are very, very similar.

[Don Davis]

Thank you very much! This is our primary box and we use them as both passive monitors and bi-amped mains. Occasionally we'll use one 712 passive over one sub per side for small gigs. These are amazing little boxes.I know you put a lot of time in on this. Again, thank you!

[jlb]

I put your data into my system for an outdoor country gig last night. RMS limiter and voltage limiters set at 500w and 110v for 712's and 1000w 120v for 718's.

I really noticed a big improvement in the performance of the 712's.

 

It is the best sounding preset that I have used so far. Thanks again Aged!

[Axisplayer]

Andy, wow. Even Don Davis likes your work, and he wrote the book. ;-)

[agedhorse]

No problem. This is a non-precision, non-lab analysis but with the experience I have doing this stuff for many years, there are some obvious solutions and approaches that were missed on this particular box by the manufacturer. It's a kick ass little box IMO, especially when exploited to it's potential.

[Don Davis]

I've never been accused of writing a book before, maybe just for using one for a pillow in class. But I used your passive settings last night with the 712's as monitors with our XTIs and was very pleased. We plan on adding a few more of these boxes to our rig.

[Axisplayer]

Don, it is just a coincidence that I find amusing.

 

http://www.rationalacoustics.com/store/books/sound-system-engineering-third-edition.html

 

It is considered a bible by many in this profession. Essential reading for anyone serious about system design.

[Don Davis]

I wish I knew that much, but it is a pretty funny coincidence. But I do learn a lot hanging out here :-)