Fulcrum Acoustic FA12, FA15 and FA28 (1 of 4)

Langston Holland

Langston Holland

Sophomore
Jan 13, 2011
222
0
16
Pensacola
Summary:

This is a review of three portable loudspeaker systems from the hand of one of the best designers in the world. Consider this a bias if you'd like, but David Gunness has a long and distinguished career with several of the finest loudspeaker manufacturers and is at the top of his game and continuing to do things that no one else is. He's also the finest technical writer I've ever read. Rich Frembes seems to be responsible for voicing decisions and these loudspeakers are the first ones I've heard in my life that I don't want to change the EQ. That's actually a little irritating, but I did find something else to monkey around with that you'll see later. :)

The FA12, FA15 and FA28 all sound the same. They are the best sounding wedges I've ever heard. The larger two maintain their sonic composure to the threshold of pain and all are extremely feedback resistant. Off axis is smooth and consistent with no surprises as you move dead on-axis with the horn. The smaller one is a bit more "open" sounding and the larger ones can go louder, both are physics things. Each loudspeaker is completely interchangeable in tone and phase. They are also light and easy to handle and have all kinds of mounting options for brackets, poles and two available angles for floor wedge use. The bracket mounts and pole cup share the same bolt pattern, thus you can move things around to taste. The NL4 cable connectors are protected by the handles on the boxes. Everything is simple and necessary - a basic definition of good engineering. Loudspeakers like this make things easy.

There are two types of processing available for these loudspeakers. One is what Fulcrum calls "Level 2" that uses conventional processing characteristics and the other is called "Level 1" that adds correction for a large portion of time domain distortion. Only a subset of the processors and processed amps on the market can import the FIR coefficients involved with the L1 presets and they do cost more. The improvement the L1 presets make is not subtle and places these boxes in a small league with the best there is.

Front:

FA_Front.JPG


Grills Removed: (FA12 horn rotated for fill use by prior user)

FA_Grills_Off.JPG


Back:

FA_Back_3.JPG


Side Panels Angle 1:

FA_Back_1.JPG


Side Panels Angle 2:

FA_Back_2.JPG


FA12 Coax Driver Front:

FA12_Driver.JPG


FA12 Coax Driver Rear:

FA12_Driver_Rear.JPG


I spent the most time working with the FA12 due to personal interest. I've been using the wonderful EONA ADRaudio M1225 self powered wedges for several years and have been very impressed with their sound quality, feedback resistance and amazingly light weight. The FA series definitely has the edge on them sonically with the L1 presets, but the M1225 is still in the neighborhood. Now that I've been spoiled with the FA wedges, I'm going to hack a flat phase preset for the M1225's' and see how much I can narrow the gap, though there is more going on with the TQ processing than flattening the phase as you'll see later.

(End Part 1 of 4)
 
Last edited:
Re: Fulcrum Acoustic FA12, FA15 and FA28 (2 of 4)

Measurements:

I made two types of measurements, one on the ground plane to simulate full range use and the other in stage monitor position that resulted in the most upward angle. A smooth cement surface was used for both, a loudspeaker to mic distance of 20' was used for the full range measurements. The loudspeaker to mic distance varied with the stage monitor measurements depending on the size of the box, but the mic was at ear position and perpendicular to the cabinet's on-axis. There was no wind, almost no environmental noise, a B&K 4007 mic was used, temperature was 14C, relative humidity was 60%. It was so perfect I measured much of my inventory until I couldn't stay awake any longer. :)

You'll notice the typical 250Hz dip in the responses in the stage monitor measurements due to a second arrival from the ground bounce that is a half wavelength late to the party. That isn't avoidable unless you made the cabinet a half sphere or bury it with its face flush to the stage surface. No one's ever bothered going to those lengths partially because that suck-out happens to sound good with a vocal mic. It can reduce the muddiness of the mix for the musician. Whatever the case, don't try to use EQ to fill in that hole because it'll dig just as deep as you boost as the woofer approaches the threshold of smoke.

Electrical:

The following two measurements are of FA12 presets with the Powersoft K3 DSP amp.

Below illustrates the complexity of the L1 relative to the standard L2 presets:

FA12_K3_L1_vs_L2.png


EQ differences between the full range and stage monitor L1 presets:

FA12_K3_L1.png


Acoustic:

Full range vs. stage monitor transfer functions of the FA12 with the Powersoft K3 DSP amp:

FA12_K3_L1_FullRng_vs_StgMon.png


Full range vs. stage monitor transfer functions of the FA15 with the Powersoft K3 DSP amp:

FA15_K3_L1_FullRng_vs_StgMon.png


Full range vs. stage monitor transfer functions of the FA28 with the Powersoft K3 DSP amp:

FA28_K3_L1_FullRng_vs_StgMon.png


Ground plane vs. stage monitor transfer functions of the M1225 self powered wedge:

M1225_Phase_GndPln_vs_StgMon.png


(End Part 2 of 4)
 
Last edited:
Re: Fulcrum Acoustic FA12, FA15 and FA28 (3 of 4)

Acoustic: (continued)

The following gets really interesting because we'll be looking more directly at the time domain differences imparted by the TQ processing. Rich was kind enough to make some L2 presets for the Powersoft K3 DSP amp for me so I could A/B the difference. The TQ processing makes a significant subjective improvement and the measurements bear that out.

Stage monitor L1 vs. L2 transfer functions for the FA12 with the Powersoft K3 DSP amp:

FA12_K3_Phase_L1_vs_L2.png


Impulse response of the FA12 L2 preset with the Powersoft K3 DSP amp:

FA12_K3_L2_IR.png


Impulse response of the FA12 L1 preset with the Powersoft K3 DSP amp:

FA12_K3_L1_IR.png


Rich also made the mistake of sending me the Lake presets for the FA loudspeakers. Only the L2 presets are available at this time since FIR coefficient import is not yet possible with these processors. This is where I started monkeying around with 2nd order all-pass filters to flatten the phase. You'll see that this improved things quite a bit, but it's still a long way from that nearly perfect TQ impulse response. There's obviously more going on with the Fulcrum processing than just phase correction.

Transfer function of the FA12 L2 Lake preset vs. the flat phase modification:

FA12_Lake_L2_vs_FP.png


Impulse response of the FA12 L2 Lake preset:

FA12_Lake_L2_IR.png


Impulse response of the FA12 L2 Lake preset with flat phase modification:

FA12_Lake_FP_IR.png


And for good measure (groan), the impulse response of the M1225 self powered wedge:

M1225_IR_StgMon.png


(End Part 3 of 4)
 
Re: Fulcrum Acoustic FA12, FA15 and FA28 (4 of 4)

The FA12 Crossover:

The crossovers in these loudspeakers are unusual enough to warrant particular attention.

These are full range "passive" loudspeakers that share some important advantages with biamped or "active" designs without requiring twice the number of processing and amplifier channels. There are no padding resistors that reduce electrical damping, waste power, reduce reliability and vary response with voice coil heating. The reduction in passive EQ L/C/R components eliminates another source of response variation with voice coil heating. Why not just use active EQ to lower the level and... drum roll please... do the EQ? The active replacement of passive processing also gives the designer a way of easing the load the amplifier will face without sacrificing SPL. And why waste amplifier output with complicated passive networks that sink current back to ground instead using it to move the drivers? There is quite a bit of this same type of thinking in EAW's KF730 passive mid/high crossover with very good results. For the life of me, I don't understand why we don't see this in far more designs since the end user will be employing a loudspeaker processor virtually without exception in the middle market and up.

I want to underscore how big a deal it is to remove the padding resistor(s) to the compression driver in a full range system. Let's look at some example specs for EAW's Microwedge 12", which from all reports is a world class product when used with the UX8800 in biamped mode. This is also a coaxial wedge, but contains a traditional "full service" passive crossover when used in full range mode.

From EAW's Spec Sheet:

Calculated Axial Output Limit (whole space SPL)

Aver - Peak
122dB 128dB (LF/HF)
123dB 129dB (LF)
128dB 134dB (HF)

Notice that when biamped, the MW12 has 6dB more headroom available from it's compression driver (1kHz and higher). That folks is a huge waste that the FA series avoids.

The Fulcrum two-way full range loudspeakers must hold the record for the most complex processing delivered through the simplest passive network. Something else you'll notice if you look closely on the circuit board is that every solder point is labeled "hand load" to confuse the competition's attempts at reverse engineering. Fulcrum keeps a special pair of 3D electromagnetic wavelength decoding glasses locked up that reveal the true solder legends that the assemblers are allowed to use only under armed guard.

FA12_Crossover.JPG


I thought it would be interesting to illustrate how this works by measuring the two passbands independently. Conceptually, the final output of these loudspeakers is a two stage process with the L2 presets and a three stage process with the L1 presets. The third stage involves TQ processing and while included in these measurements is not relevant to the discussion.

The first stage is the acoustic output of the passive network and driver after the amplifier. The second stage is the electronic processing before the amplifier. The raw driver overlap between passbands is largely dealt with in the first stage and when combined with the active processing delivers the intended response. To do this well, this type of design still requires a rather beefy compression driver and physically close acoustic origins between passbands. That means an expensive coax driver with a horn that transitions well into the woofer's polar response throughout the crossover region.

Measurement Setup:

FA12WedgeMic.JPG


I used Smaart v5.4 for the following plots due to its coherence blanking feature that hides the useless out of band phase noise. If you study these closely you'll see that they agree flawlessly with the previous full system measurements using the much more detailed CLIO even though they were made during different measurement sessions. I really need to get with the program (groan) and learn Smaart v7, just been too busy and CLIO stole my heart.

The following measurements show the FA12 acoustic woofer output(1) in stage monitor position with the passive filter alone (cyan) and with the addition of the active filter (green):

FA12_Act_Pass_Low.png


The following measurements show the FA12 acoustic compression driver output in stage monitor position with the passive filter alone (blue) and with the addition of the active filter (purple):

FA12_Act_Pass_High.png


Now we see the active plus passive transfer functions of both woofer and compression drivers with complimentary phase curves:

FA12_Act_Pass_Both.png


Footnote:

1. The grill was removed for these measurements to make it easier to disconnect and reconnect the high and low passbands as needed. The unused passive crossover leads were connected to a 10Ω power resistor. I also shorted the positive and negative driver terminals of the unused passband. See below.

FA12Internal.JPG


(End Part 4 of 4)
 
Last edited:
Re: Fulcrum Acoustic FA12, FA15 and FA28 (4 of 4)

Wonderful as usual Lang. I'd love to have a small fleet of these, but sadly they are a bit beyond budget for me.

Nice job by the way on the "flat phase" Lake preset you made.

I am a big fan of Fulcrum's idea of implementing a simple passive network to allow for single channel use combined with substantial DSP. It adds significant economic value and terms of DSP and amp channels, and reduces complexity without negatively impacting performance.

A question: what are those yellow dots on the horns?

Cheers
Jeff
 
Re: Fulcrum Acoustic FA12, FA15 and FA28 (4 of 4)

Jeff said:
A question: what are those yellow dots on the horns?
Click to expand...

I'm under a non-disclosure agreement on that.

But, given my little brother that works in the ad dept at NEXO blabbed about the CL3 a day before he was supposed to... They are Italian wine bottle cork fillers for the bolt holes that hold those custom horns to the B&C driver. I'm not dead sure about the cork source, but I suspect it after my trip to the NY AES convention. :)
 
Last edited:
Re: Fulcrum Acoustic FA12, FA15 and FA28 (4 of 4)

Hey, Langston. Thanks for the thorough review. I can't recall seeing gear so carefully investigated and objectively reported anytime recently. I also found your Powersoft K series review to be terrificly enlightening.

To all those who may be interested - stop by the Fulcrum stand at InfoComm and we'll reveal the mystery of the corks. We'll also have the crossover decoder goggles available for anyone who wants to try them ;)
 
Re: Fulcrum Acoustic FA12, FA15 and FA28 (4 of 4)

I had the good fortune of taking 6 FA12's on tour with me last fall for Joan Baez and Kris Kristofferson. I'm over in the UK finishing up 5 weeks with D&B Max 12's, and feel like I'm slumming it. As I've said before, the FA12 is the best wedge I have ever used. The only issue I had was the latency caused by the FIR. Coupled with the latency of the SC48, it created an audible delay. None of the musicians noticed. Somehow the LD was the first person to mention it. He's also a drummer. He has been sacked.

Langston, did you measure the latency using Powersoft K3 amps? I had Itech HD's out on the tour. I imagine that when the Fulcrum DSP is released the latency will be down to an unnoticeable amount. If I were doing it again I would also get an AES output card for the console to avoid the latency of one DA and one AD conversion.

Does anyone have any in the UK or Europe? I have a 7 week tour in June and July and would love to get back on the FA12's. I've got L'acoustics 115 FM's reserved, but would prefer the Fulcrum boxes. Unfortunately shipping costs prohibit sending them over.
 
Re: Fulcrum Acoustic FA12, FA15 and FA28 (4 of 4)

Jason said:
... did you measure the latency using Powersoft K3 amps? I had Itech HD's out on the tour. I imagine that when the Fulcrum DSP is released the latency will be down to an unnoticeable amount. If I were doing it again I would also get an AES output card for the console to avoid the latency of one DA and one AD conversion.
Click to expand...

Hi Jason:

Your observations support a theory I've had about the latency issue, but I don't have enough experience to nail it down. I cannot hear any latency issues with the FA series driven by the Powersoft DSP amps. The Powersoft amps have about 1/2 the latency of the ITech HD's with the Fulcrum FIR coefficient presets, but this shouldn't be necessary and hopefully Crown can provide a fix soon. While the Powersoft amps are native 96kHz animals, the HD's can switch between 48kHz and 96kHz, the latter cutting latency in half - but that's not the end of the story.

I measured about 5ms latency in the Powersoft amps. The Fulcrum L1 presets only add about 1.7ms to that since no FIR crossover filters are involved. It requires at least 5ms additional delay to implement usable FIR crossover filters, but not so for time correction such as done by the Fulcrum process (or All-Pass filters BTW). The difficulty with the HD amps that I learned from Rich is that it appears Crown apportioned an additional 5ms or so crossover latency into the amp when importing FIR coefficients even if no crossovers are involved.

Whoops.

I'd be very surprised if someone could detect the latency with a Powersoft DSP amp using the Fulcrum L1 presets.

---

On the idea about eliminating the D/A/D process from console to amps by driving them digitally, I'm confident that would help with a console that does 96kHz natively. I don't have the knowledge to comment on what the latency issues would be if the console runs at 48kHz with the Powersoft DSP amps converting to 96kHz internally. I know Powersoft knows the answer to this. :)
 
Re: Fulcrum Acoustic FA12, FA15 and FA28 (4 of 4)

Just for grins, did you measure the combined passive response without the DSP eq upstream? Not that this would EVER happen.
 
Re: Fulcrum Acoustic FA12, FA15 and FA28 (4 of 4)

Hi Brian:

Actually, all acoustic measurements I do these days are recordings. I use Smaart v5.4 to monitor the recording to be sure I'm getting good data and use Audition v(latest) to record a stereo file to disk. Left is the measurement through the mic and right is a loopback. I make sure I have more than 10 seconds of good data before I end the recording, then I trim the thing down to exactly 10 seconds and put a read-only attribute on it. I also record temp, humidity(1) and distance(2) from mic capsule edge to front grill edge.

That means I can go back at any time in the future and loop that recording through a processor and re-tune, re-align or in some cases re-member what I did. :)

I can loop these recordings through any two channel FFT software out there - that's almost all of them. If the software only accepts an IR, I can make an IR out of the recording in ARTA or Smaart.

IOW, what you ask is easy. I already saved those FA12 traces in Smaart during the recording monitoring process, so that's what's shown. It's also easier to compare with the independent passband measurements in part (4 of 4) above due to the use of Smaart there as well. The passive crossovers in these loudspeakers were quite functional prior to the active processing. A "judicious" use of EQ could get you through a gig with these wedges if your processing device broke.

Cyan trace is in the ground plane at 20', Green trace is stage monitor position with mic at ear level:

FA12_Passive_Only.png


Footnotes:

1. Fluke 971 Temperature Humidity Meter

2. Leica DISTO D5
 
Last edited:
Lang,

I wonder if you need a different preset in the case of running those amps at different sample rates? Due to Nyquist error most DSPs behave differently above 10kHz depending on sample rate.

P.S. I find that Leica D5 to be the second most valuable item in my kit, after Smaart. For temp/humidity I use a Kestrel 4000, while farther down that list it is a great asset and runs a long time on standard batteries. Also fun to take on hikes.
 
Re: Fulcrum Acoustic FA12, FA15 and FA28 (1 of 4)

First off - kudos to Langston for his usual excellent analysis!

The latency associated with A/D and D/A is primarily due to the anti-aliasing filters. In today's high quality A/Ds, it is typical to oversample with a low-order analog anti-alias filter, then use an FIR-based brickwall, linear phase low-pass filter as the anti-alias filter before downsampling to 48 kHz. The linear phase part produces about 0.5 ms of latency on both the input and output, so between the A/D and D/A there is about 1 ms of latency. To that is added firmware latency (i.e., DSP calculations are often done in "frames" of 8 or more cycles, because the coefficients can be loaded once and applied multiple times instead of once). You could realize the same anti-alias filters with half the latency at 96 k Hz, but since the filters don't need to be as sharp, you can use shorter FIRs and reduce it even further.

We're currently in a (temporary, I'm sure) situation where the digital electronics industry hasn't kept up with what us speaker guys are doing with processing, and they're guessing to some extent. "Why are we putting this FIR in here?" "Umm. I think the speaker guys want brick wall crossovers." "OK - well then we should put in a compensation delay set to the latency of the lowest frequency brickwall filter they might use. That way, we can tie it to the latency of the filter they select so the latency doesn't change if they change the frequency." Thanks guys for thinking of us, but no thanks. They will eventually figure out the difference between loudspeaker designers creating presets for their products and users getting confused when the arrival time changes because they changed a crossover frequency. We'll get there, but in the mean time, we'll have to deal with things like an I-Tech amp that puts in 5 ms of unnecessary latency in my loudspeaker settings, supposedly to save you all from getting confused ;-)

Oh, and Bennett: You definitely need a different preset when running at 96 kHz, regardless of the high frequency response of IIR filters. FIR filters are essentially impulse responses, so if you change the time scale from 20.6 us to 41.2 us everything shifts up an octave in frequency. I can tell you from experience that that doesn't sound very good ;-)

Dave
 
Re: Fulcrum Acoustic FA12, FA15 and FA28 (1 of 4)

David Gunness said:
Bennett: You definitely need a different preset when running at 96 kHz, regardless of the high frequency response of IIR filters. FIR filters are essentially impulse responses, so if you change the time scale from 20.6 us to 41.2 us everything shifts up an octave in frequency. I can tell you from experience that that doesn't sound very good
Click to expand...

Oh, of course! Funny, that. I figured you guys would have been handing out different presets anyway, but was curious. Can you tag the presets so that they'll only run at the right sample rate?
 
Re: Fulcrum Acoustic FA12, FA15 and FA28 (1 of 4)

Some platforms include a "sample rate" tag, but most don't, and the ones that do don't necessarily prevent a wrong sample rate from being used; so Rich has to tag the files as "48 kHz" or "96 kHz". We'll eventually get to this issue in the AES standards committees; when it does, I'll advocate a standard AES format for FIR files that includes sample rate and/or a standardized format that can be converted to any sample rate. But for now, folks just have to be careful with what they're loading.

D
 
Re: Fulcrum Acoustic FA12, FA15 and FA28 (1 of 4)

Something to chew on...

Powersoft amps have 5.02 ms latency when using analog inputs and with no filters applied. I see their latest firmware update bumps this latency up to ~6 ms when AES3 inputs are used; analog inputs remain the same. This throws a bit of a monkey wrench into Jason's idea of using AES3 inputs to reduce latency. :?~:-?~:???: Loading, for example, our CX896 preset adds 1.56 ms latency.

Crown I-Tech HD amps, using analog inputs and no filters, have 1.02 ms latency at 96 kHz SR and 1.66 ms at 48 kHz SR. Loading the aforementioned CX896 preset adds 3.72 ms at 96 kHz and 2.39 ms at 48 kHz. (I used the CX896 preset example because I happened to have a 48k Crown preset readily available).

Bennett Prescott said:
Can you tag the presets so that they'll only run at the right sample rate?
Click to expand...

The I-Tech HD amps save the sample rate with a preset. That was smart of them. We have seen a customer get burned by using a 48k FIR in a popular drag-and-drop installation processor that was set to 96k SR. They couldn't understand why we would publish a preset that sounded and measured so poorly. Bet they won't make that mistake again! Still, to Dave's point it would be nice if there was a warning that something was amiss.
 
Re: Fulcrum Acoustic FA12, FA15 and FA28 (4 of 4)

Langston Holland said:
Actually, all acoustic measurements I do these days are recordings. I use Smaart v5.4 to monitor the recording to be sure I'm getting good data and use Audition v(latest) to record a stereo file to disk. Left is the measurement through the mic and right is a loopback. I make sure I have more than 10 seconds of good data before I end the recording, then I trim the thing down to exactly 10 seconds and put a read-only attribute on it. I also record temp, humidity(1) and distance(2) from mic capsule edge to front grill edge.

That means I can go back at any time in the future and loop that recording through a processor and re-tune, re-align or in some cases re-member what I did. :)

I can loop these recordings through any two channel FFT software out there - that's almost all of them. If the software only accepts an IR, I can make an IR out of the recording in ARTA or Smaart.
Click to expand...


GREAT idea. These are the types of things I see and say to myself - "why didn't I think of that!?!"
You are a real treasure in this community Langston!

Great thread!
 
Top Bottom