OT12's with compression drivers

[gdougherty]

It's been about a year since I swapped out melded arrays in 6 of my 8 OT12's for compression drivers and apparently I never wrote a review.

I've had about everything except for a old style crossfiring tweeter array in OT12's now. I started off with straight arrays in my first OT12 builds and never was really happy with them. They just didn't keep up and sounded harsh and brittle when pushed hard. At the time I was running a total of 4 OT12's and the mid-bass had more gas but the tweeters frequently disappointed. Not to mention, there's that hole at 1.2-1.3K I've always had from baffle reflections. Newer models may have fixed it, but both my builds suffered from it and the piezos left it hanging out there.

Trading up to melded arrays definitely improved things. I had plenty of output under most circumstances, wide dispersion and the arrays sounded much better. However, they were not without their limitations. The horizontal dispersion I'd describe as too wide. The mid-bass couldn't keep up at the outer limits of the HF coverage so you ended up with a low-mid and 2K+ focused response out to the sides that wasn't particularly pleasant, though it might be clear. Tinny is a good adjective. While output capability was improved, I still wasn't happy when I pushed the boxes. They still fell apart and had a bit of an icepick feel to them. On top of all that, the diffusion rod in the middle helped things some, but there's still a very noticeable beam directly on-axis to the cabinet caused by the crossfiring elements in the melded array.

So, that's the limitations I saw, but there are very good financial reasons I went with the piezos to start, and kept using them for several years. Piezos cost me about $50-60/box selecting 14 of the best I could find after ordering about 30% more than the build required. The passive HP filter cost all of $14 if I remember correctly so along with about $5 of wire and a $4 bottle of PVC cement I built everything I needed to have complete HF arrays. Buy completed arrays from Leeland like I built for all of $112 or standard 12 element arrays for just under $100. On top of that, you can easily drive 4-6 OT12 off both channels of a single higher power amp.

The comparison cost? I spent a while looking for a driver that would crossover at 1.2K or lower to help cover the hole in the LF response and would have enough output to keep up with how I often push my OT12's. I settled on the Beyma CD10Fe. There are other options, but I like the non-metallic diaphragm sound and many of the other features for a very acceptable price. Shipped and everything, I think I got mine for about $140/ea. I had no time or patience for building the custom horns, and Greg Plouvier (God rest his soul) agreed to put together 6 of them for $300.

So for about $1100 on top of the ~$250/cab I originally spent to build my OT12's, I was just about set. That is, until I decided to skip the passive crossover and go biamped so I could align the HF and LF sections in DSP. In the budget range, the newer XTi2 models were just the ticket due to their nifty limiting which is adjustable by output voltage rather than the input signal. A 4002 and 2002 (with another set to someday follow) added about $1600 to the total upgrade cost and freed up the plx3602 I used to run on tops for sub duty. So, a few hours of painting, sanding, demo work on the old melded array supports and $2700 later, we were ready to rock. Oh yeah, there was also the 250ft spool of 12/4 and all the nl4fx connections to make enough of various lengths to run two biamped channels on each side of the stage and the updated i/o panel on the amp rack for my new connections. Call it about $3100.

The question that follows is of course, "Was it worth it?" Well, that depends on what you value I guess. For me, it was a very worthwhile trade up. I've pushed them hard and run the entire rig with my leftover melded arrays serving as center fills and been extremely happy. The compression drivers are nice and smooth at all volumes and I no longer get that icepick to the brain sensation when I crank it up loud. The dispersion of the custom horn seems to match the LF dispersion of the box better and the mids are much better now that the 1.2K hole is mostly covered up. The HF extension is now also improved since one side effect of the melded array seems to be a rolloff above 12K or thereabouts. Dispersion isn't too wide up there anyway, but it is there now without needing a big EQ boost that introduces distortion in the melded arrays if it happens to clip in your signal chain.

[Grant Bunter]

Thanks for the run down on thoughts and costs and the review!

Do you think there's enough coverage of the 1.2k ish hole you mention, now covered by the CD's, to allow a pair of them on top of a melded array pair either side?

[Bas Gooiker]

Thanks for this very extensive review.

It sums up about all my fears i had about using piezo's. I have been considering those Beyma's as they are really good for a budget (75 euro here) CD driver. Did you remove the crossover networks completely when you started by amping?

[gdougherty]

Thanks for the run down on thoughts and costs and the review!

Do you think there's enough coverage of the 1.2k ish hole you mention, now covered by the CD's, to allow a pair of them on top of a melded array pair either side?

Mine have more of a notch than a hole. It's centered around 1.26K and it's only about .2 octave wide. Most people don't notice it, but I did when I compared the stack to my reference headphones and it always showed up on measurements. Putting a CD over a melded would fill some of that in, but it's not anything I would lose tons of sleep over. It would also be something I'd be leery of since you introduce another timing variance into the stack and the net result may be worse off than running one or the other. The CD horn is 5-6" deep where the melded array is more like 3-4". If both cabs were biamped and you had independent delay on them, you could probably straighten it all out but I wouldn't recommend going there otherwise. The horizontal on the custom horn is ~100 degrees so you're not losing much either way.

I left my last pair as melded because 3 per channel was the most I cared to subject the amps to and I had another 2 channel amp I already used for center fill or other purposes, so it made sense to keep the last pair melded until I was ready to biamp or replace those.

[leamy_f]

Yes thank you for this review. Could you let us know what settings you applied to your crossover in respect of both the lf driver and the comp unit in terms of frequency and slope and also the voltage limit you applied to the beyma's....regards..Frank Leamy

[gdougherty]

Thanks for this very extensive review. :clap:

It sums up about all my fears i had about using piezo's. I have been considering those Beyma's as they are really good for a budget (75 euro here) CD driver. Did you remove the crossover networks completely when you started by amping?

If you have the budget for CD's and doing them properly, it's not a bad way to go. If you don't, the piezos aren't a bad option for most users around here. Most users being smaller bands playing in smaller venues with modest volume needs. Those situations never gave me a problem with the melded arrays. The ones where I really noticed were when I pushed things just a little harder to get ~105+db at FOH in a venue that holds 700-1000 people. For my own hearing I don't care to do that, but I wasn't always the one in charge and I let them deal with the results.

The dispersion problems were most notable outdoors with no boundary reflection of the LF range and where the coverage area was wide enough to allow you to really walk out to the edge of the speaker's coverage.

Indoors I often deployed the speakers wide and toed them in a little bit to do as much as I could for stereo imaging in the middle of the speakers. Being between the speakers always provided a better full-range experience where the LF pattern overlapped between the two. Also note that I'm running only 3-4 per side. Other users reporting great results with the melded arrays have used 5-6 boxes per side to cover similar or larger crowds. Realistically, that would have been a better option though in the past I was limited by pack space, amp channels and time to build another 4 speakers.

I did also find that the OT12's aren't the best at stereo imaging since the ports create a more diffuse sound than you'd get with a direct radiator. I put my ports on the outside edge of the back panel which probably only exacerbated the diffusion effect. Listening to them side by side with a pair of Apogee AE5's the AE5 was very easy to audibly locate, while the OT12 sounded much bigger and less specific in its position.

I followed the plans on the crossover network for the CD filters. I left out the HP for the HF section since it has no effect other than dropping the lows. The LP filter on the LF driver however helps in the sensitivity and since it's centered up around 2K didn't seem like a bad thing to leave in. The plans also suggest it as an option IIRC.

[gdougherty]

Yes thank you for this review. Could you let us know what settings you applied to your crossover in respect of both the lf driver and the comp unit in terms of frequency and slope and also the voltage limit you applied to the beyma's....regards..Frank Leamy

1.2K center Linkwitz Riley 24db/oct for the crossover. Forget the value, but I used the RMS specs on the Beyma driver along with Ohm's Law calcs to find a voltage limit for the HF section. That would be the safest setting. You could also opt to limit using a program rating or 2x RMS if you know you've got enough output and will treat things well.

[osse]

Interesting indeed, now I wonder if any of those problems applies to the DR line too? could you expect the melded array to have too much dispersion?

[Bill Fitzmaurice]

I did also find that the OT12's aren't the best at stereo imaging since the ports create a more diffuse sound than you'd get with a direct radiator. I put my ports on the outside edge of the back panel which probably only exacerbated the diffusion effect. .

The output of the ports is centered at 100Hz, extends only to 200Hz, and is completely omni-directional, so it doesn't matter where they're placed. For that matter the output of the front wave is also omni-directional up to 200Hz, as it is with any cab less than about two feet wide.
The wider dispersion of the melded array than the upper range of the woofer would be most noticeable outdoors at the far edge of the soundfield, mainly because you're able to stand there. Between room reflections and the general inability to stand at a 90 degree or more angle relative to the cabs it's seldom noticeable indoors.

[gdougherty]

I did also find that the OT12's aren't the best at stereo imaging since the ports create a more diffuse sound than you'd get with a direct radiator. I put my ports on the outside edge of the back panel which probably only exacerbated the diffusion effect. .

The output of the ports is centered at 100Hz, extends only to 200Hz, and is completely omni-directional, so it doesn't matter where they're placed. For that matter the output of the front wave is also omni-directional up to 200Hz, as it is with any cab less than about two feet wide.
The wider dispersion of the melded array than the upper range of the woofer would be most noticeable outdoors at the far edge of the soundfield, mainly because you're able to stand there. Between room reflections and the general inability to stand at a 90 degree or more angle relative to the cabs it's seldom noticeable indoors.

Maybe its a result of the rear ports I did and despite the wide spacing, still getting some upper frequencies off the back of the driver as well as the wide dispersion of the melded array reflecting off side walls. Flown about 2ft apart a pair of direct radiator cabs was very direct and the OT12 was more diffuse which reduced the apparent stereo imaging as you moved across the sound field. Most listeners wouldn't do that or notice but it was a noticeable difference.

And yes, as we both noted the dispersion on the melded array was most noticeable outdoors. I'd hear it to some extent in wide rooms as well but it was less off a problem. As noted though, the compression driver horn is a better match and doesn't exhibit that at all.

[gdougherty]

Interesting indeed, now I wonder if any of those problems applies to the DR line too? could you expect the melded array to have too much dispersion?

On the dr280/90 it could possibly be even more noticeable since the bigger boxes narrow the pattern of the LF horn. That said, by all reports the 280 with a 3012lf and a pair of top quality compression drivers is a killer system that will cover several thousand with 3-4 boxes per side. If I had the gear and engineering chops to fly a rig like that, its where I'd be.

[Grant Bunter]

It would also be something I'd be leery of since you introduce another timing variance into the stack and the net result may be worse off than running one or the other. The CD horn is 5-6" deep where the melded array is more like 3-4". If both cabs were biamped and you had independent delay on them, you could probably straighten it all out but I wouldn't recommend going there otherwise.

I'm interested on your thinking here as well, so I'd appreciate it if you, or anyone else, would share thoughts on the following:

From what I understand, in distance vs delay, it ballparks that for every foot you delay by 1 millisecond. So as an eg, if you were matching back line to FOH and the backline was 15 feet behind the FOH, you delay the FOH by 15 milliseconds.
I know some here also delay tops/mains and subs based on the horn path length of their subs. I don't, I think mids and HF a tad in front of LF is fine.
Now you're suggesting one should consider using delay to match mids/HF for a 2-3 inch mismatch in path length???
Does that mean you delay everything, always, adjusting for the setup on the day?

As I said, love to hear the thinking from people who work this way...

[Bill Fitzmaurice]

Now you're suggesting one should consider using delay to match mids/HF for a 2-3 inch mismatch in path length???

If you have the ability to do so it won't hurt, but I wouldn't jump through any hoops to do it.

[Grant Bunter]

Thanks Bill!
If the DCX is capable of less than 1 millisecond delay increments, yeah I could do it, but I really wonder if it's even noticeable.
I'll check that out...

[gdougherty]

Thanks Bill!
If the DCX is capable of less than 1 millisecond delay increments, yeah I could do it, but I really wonder if it's even noticeable.
I'll check that out...

It's less of a problem for an individual speaker because outside of the crossover region there is no interaction. When you get to multiple speakers you can introduce comb filtering depending on the spacing. Whether or not to do it depends on how far you want to chase optimal sound. The results will be better with the timing of everything aligned but very usable sound can be had without worrying about it.

I worry about all that enough that I use the HF section as the axis if I rotate one cab on top of another. I get enough strange looks from the DIY cabs in some circles that I'd prefer to always present them in the best light.

Did a benefit last night in an event center and had numerous compliments on the sound. Had a manager comment that they'd seen many other large systems roll through and was amazed at how good things sounded with the relatively small number of speakers I brought. Had 3 OT12 w/ CD per side up on my double stand platforms and the melded arrays on the 16ft line of subs against the front of the stage as front fills.