DR300 response

[snowphish]

We built 6x DR300s a few years ago, with Eminence Impero 12s and NSD 2005 drivers - BI-AMPED.

For the longest time I would just let the Driverack PA2 do its autoEQ - and not think much of it because I don't have a space or my own to do any tuning, and there's usually not enough time before gigs to tinker.

I always noticed there's a gap around 1k on my tops that the autoEQ does its best to adjust for - I always figured this was due to the cross-over point phasing issues and while I did try playing around with my curves, I never got around to measuring and finding the root problem and figured it sounded GOOD ENOUGH with just the driverack's autoEQ.

Recently we've been getting festival gigs, and those allow me to have more time to play around and tune it manually, time/phase align the tops with the subs with SMAART, play around more with the EQ.

I decided to take some measurements, one side only - extending the cross-over beyond the usual ~2k cross-over points and measuring mids/highs independently. I am realizing that the mids fall off around 1k, and the highs fall off around 2k - so that explains my huge gap between 1-2k - so it appears that it isn't caused by phasing issues at my cross-over point.

I know the NSD2005s are usable down to 1.5k - but the IMPERO 12s should be good well above 1k.

In order to get a more linear response I have to EQ almost +12db around the gap for the mids - I had a manual EQ for the last gig we did to flatten things out as much as possible, rig ran for 116 hours, international headliners and all, and everyone thought it sounded good.

What could be causing the drop-off I'm seeing? Is +12db really a reasonable amount to fix that issue?

Pic of the rig

[Bruce Weldy]

The NSDs can be crossed as low as 1.2 with the proper slope. My OT12s are running Bill's passive crossover at 1.2k and if anything, I end up cutting a bit between 1-2k for a smoother vocal sound.

Any chance that there's something in another piece of gear that has a big cut in that area? You sound like you know what you are doing, but have you checked the mixer's main EQ to make sure something isn't amiss with the graphic or PEQ on the main bus?

I guess the ultimate question is - does it sound better when you add 12db in that area? Or, are you strictly looking for the graph to be flat?

[Bill Fitzmaurice]

You should be crossing over at 1.2kHz.

[snowphish]

You should be crossing over at 1.2kHz.

When I am measuring the HF with mids muted I have the crossover lower than that and they still roll off around 2k .

I will play around some more next time I have the chance and take some screen shots.

[Bill Fitzmaurice]

Something in your signal chain isn't right. There's also the possibility that the HF drivers are wired out of polarity.

[Seth]

The rig's looking pretty good, Andy

When I am measuring the HF with mids muted I have the crossover lower than that and they still roll off around 2k .

- What specific frequency was the high pass on the NSD's set for?
- What slope?
- Bessel, Linkwitz-Riley, Butterworth?

When you say "they still roll off around 2k", do you mean they start to roll off on a measured chart or is it something you hear?

[mj90210]

I have the exact same issue with my DR300's using impero 12A and NSD2005's using the plan crossover and the even biamped. I got a 3-4 db extra by time aligning when biamped. I loaded the DR300's with B&C 12ndl88 and B&CDE250 and have similar results. I am no master with measuring but it was done with REW while the cabinets were elevated, away from boundaries outside.

[snowphish]

- What specific frequency was the high pass on the NSD's set for?
- What slope?
- Bessel, Linkwitz-Riley, Butterworth?

When you say "they still roll off around 2k", do you mean they start to roll off on a measured chart or is it something you hear?

I was measuring this with SMAART, on axis, a few meters away from the tower with a dbx RTA mic. Not sure what my slope was, likely LR24 or something like that, but for the measurement I set the crossover point below 1k and muted the mids. The highs rolled off around 2k, pretty steep roll off too.

I think I ended up crossing over around 2k and EQing the mids to fill that gap.

I guess I will try to get some measurements of individual cabs to eliminate potential issues between the 6 CDs in one stack...

[Bruce Weldy]

There are a lot of places in the signal chain that could be having an effect on those frequencies. Check the mixer for EQ on the mains. Check the driverack for any PEQ EQ settings that are outside of the Graphic EQ settings. Recheck the Crossover settings.

And it wouldn't hurt to fire up a new Driverack setting from scratch and rebuild your crossover points, etc.

[mj90210]

I have measured mine 100 different ways, woofer only to compression driver only with the crossover and without. I am 1000% sure it is not in the signal chain(s).

[Bill Fitzmaurice]

I have NSDs in my OTops, they don't show a roll off below 2kHz. However, they do have a rising sensitivity as the frequency goes higher. To account for that I EQ the high end down. Anytime you EQ you want to use an equal amount of boost and cut, not just one or the other, to get the desired result.

[mj90210]

Bill, when measuring the woofer and compression drivers separate in the cabinet, you can see the woofer accounts for the large dip. I even changed the dampening and materials in the cab to see if it would help. I have triple checked the throat filler. As I previously stated I have tried completely different drivers and a I still have the same results as before and the original poster.

[Bill Fitzmaurice]

The woofer can't account for a dip that occurs above its passband.

[Seth]

I was measuring this with SMAART, on axis, a few meters away from the tower with a dbx RTA mic. Not sure what my slope was, likely LR24 or something like that, but for the measurement I set the crossover point below 1k and muted the mids. The highs rolled off around 2k, pretty steep roll off too.

Good info to work with

You may already be aware, but I'll put it out there anyways... depending on the filter type, the signal is already attenuated -3dB to -6dB at the designated cutoff frequency. A LR24 set to 1000Hz starts rolling off upwards of 2000Hz, and is down -6dB at 1000Hz. So, if by "below 1k" you mean something like 950Hz, the filter is still greatly effecting your 1k-2k response measurement. If by "below 1k" you mean something like 350Hz, that would give us a better idea of what the response of the driver/horn combo actually is in the 1-2k range.

I guess I will try to get some measurements of individual cabs to eliminate potential issues between the 6 CDs in one stack...

That sounds like a good plan. I'd be interested in seeing the test results. It's also a good idea to test each portion of each cab individually to check uniformity among the herd and also each portion of each cab together with a known good cab to verify all the cabs are wired the same and are all in phase too.

The most common cause for attenuation in frequencies near the crossover point is simply having the mid's and highs out of phase. I'm sure it's something you've looked into, but it's worth revisiting and starting with basics.

[Seth]

In rereading the initial post, it looks as though you had your system setup incorrectly from the get-go. If you don't have time and space to do all the testing prior to your next gig, set the crossover point at 1.2k on a LR18 and see if that improves things.