[ekovalsky]

I've been working my ass off this week and have only had a few hours to play with the TacT and the new speakers.  Here is the latest measurement curve, after moving the speakers around a little.  Remember these are in-room measurements at the listening position, not nearfield or quasi-anechoic measurements like you see in manufacturer specs or magazine reviews.

I am running the main channels full range, as designed.  There are room suckouts at 54hz & 110hz.  Later this year the room will be expanded to approximately 17' x 23' and I will be getting professional advice on exact dimensions and treatments to keep nulls at a minimum.

Using TACS I implemented a 3rd order butterworth filter (-3dB @ 50hz) on the S2150 driving the subs, along with two bands of ParEQ to fix the 39hz and 66hz room nodes.

So far, none of my attempts at correction have resulted in any improvement of the sound.  Thus I am listening in the bypass mode.  This is in contrast to my previous speakers which sounded much better with correction implemented compared to bypass.  Anyway, for a 'flawed' design using "terrible" metal cone woofers, "cheap" alnico midranges, and too short of a line array these Alons measure pretty damn well.  And the sound, even in my suboptimal room with an amp out of its element driving the sub towers, is up there with the best I've heard anywhere -- imagine the accuracy of headphones, but with an incredibly wide and deep soundstage and the dynamics of a live performance

I expect more improvement with the room expansion and addition of a brute solid state amp comfortable with 1-2 ohm loads on the sub towers.

[Bingenito]

Congratulations

Not requiring room correction is interesting. When I had the arrays I used a DBX DriveRack (not TACT or DEQX but it does room correction).

As much as I tried to tweak with the correction in the DBX the sound was always best without it. The only exception to this was using the room correction for the Peerless XLS subs.

I am willing to bet that the larger room will make a positive improvement.

[ctviggen]

I take it that the blue and orange-ish curves are prior to the TACT?  What about after the TACT?  Does the TACT allow you to assign the curve you want, and also show the resonse with the TACT in the loop.  I pulled up my ETF plot of my RM40s, but it's a bit hard to compare -- ETF uses absolute dB readings.  I'll try and post my readings when I get time.

Anyway, it doesn't sound boomy uncorrected?  From your curves, if you add the sub towers and the reg towers, the below 500Hz seems as if it'd be very high.

[hometheaterdoc]

I take it that the blue and orange-ish curves are prior to the TACT?  What about after the TACT?  Does the TACT allow you to assign the curve you want, and also show the resonse with the TACT in the loop.  I pulled up my ETF plot of my RM40s, but it's a bit hard to compare -- ETF uses absolute dB readings.  I'll try and post my readings when I get time.

Anyway, it doesn't sound boomy uncorrected?  From your curves, if you add the sub towers and the reg towers, the below 500Hz seems as if it'd be very high.

Look at the frequency plot picture.  Just below the 100Hz point on the plot image, you should see the word View and then L, R, Sub-L, Sub-r, and CRO, each with different colors.  Each of those colors corresponds to a frequency response measurement on the plot.  

The frequency response plot you see for each channel is the raw, uncorrected measurements.  

The neon green line is the target curve that you want to achieve after Tact applies its correction.  The red dots can be moved around to create any kind of curve you desire.  You can add more points to the target curve line as well in case you want to have very narrow adjustments.  When you have the curve you want, you calculate the correction curve and program it to one of the 9 seperate correction memories.  The software will go in 1Hz increments and map where the measurement point was and bring it up or down to where the target curve point is.  What is also done, and what isn't shown on the crude graph, is time alignment and phase alignment so you end up with an incredibly coherent presentation with great spatial qualities.

This setup has no crossover defined (as can be seen under the crossover section with LP_None.CRO and HP_None.CRO).  But you can put up to a 10th order (60dB per octave) filter between 50Hz and several hundred Hz.

In this case, if I am reading the original posters comment right, since he is using Tacts own digital amps, you can program a crossover in the amp itself and don't necessarily have to do it in the 2.2X preamp.

ekovalsky:  these do measure reasonably well for an in-room measurement.  I've seen better measurements on a few models I've measured.  But I would hazard a guess they don't sound as good as these do.  the new Dancer II CP-8571s from Usher look like someone took a ruler and drew a straight line from 2K through ~13K.  It freaked me out when I first measured them

I'm glad you're enjoying the new speakers.  I'll be curious to see if after more tweaking you still prefer no correction at all.

[ekovalsky]

hometheaterdoc,

You description of the TacT screen and setup is entirely correct.  There is free software called "TACS" available from Dirac in France which lets you create customized filters for the TacT amps which can incorporate any of the following:  crossovers of any slope/frequency, parametric EQ and tone control, delay/phase adjustment, level adjustment, and room correction.  This gives much more flexibility over what the RCS alone can do and also makes time alignment between different driver groups (in my case just mains and subs) much more accurate.

The Usher speakers seem to be quite interesting and probably underpriced in the current market.  Do they use some higher order passive crossovers, which typically make measurements much better at the expensive of phase integrity ?  The Alons's external passive crossover is all first order or quasi first order, with all the driver groups in the same polarity with phase coherence.

At some point I will probably bypass the external crossover box and multi-amp the mains with FIR higher order DSP filters.  Since the speaker already has a pair of binding posts dedicated to each driver group, with no intervening circuitry, it practically begs for this.   It will definitely measure better if I do that, but I am no longer convinced it will improve the sound.  Time will tell!

[JoshK]

I'll go out on a limb and hazard the guess that you don't have suckouts at 54hz & 100hz but rather have gain just below those frequencies.  I.E. the valley is more the anechoic response tapering off slowly and there are four gain related humps from the room.  ~40hz, 80hz, 160hz, 240hz.  This is a guess but I'd bet a small sum on it.   If you were to bring these humps down with traps or whatever else then you sub towers would compensate for the slow slope from the main towers to give a flat sum.

[JoshK]

If the Alon's use first order xo's then how does he control the nasty break-up modes of the SEAS and presumeably other drivers?  The Excels are the W26's (just by visual guess)?  

The W26 isn't a magnesium cone, it is an aluminum cone unlike the rest of the Excel lineup, it is merely painted to match the cosmetics of the rest of the Excels.  This isn't necessarily to say that it is a bad driver, just that it isn't mag, its aluminum so the breakup are going to be even worse than if it were magnesium.

[ekovalsky]

The crossover filters of the Alons are rather complex, and I think the woofers have a quasi second order low pass crossover.  The first breakup of the Seas is at 5khz (according to Carl M.) and with the crossover at about 400hz, output is down enough at 5khz to not be a problem.

If anyone wants I can post pics of the crossover internals when I get home from work.  When opened, there are three separate circuit boards in each, one for each driver group of the main channel.

[JoshK]

It is probably low order plus a notch filter.

[Rick Craig]

I've been working my ass off this week and have only had a few hours to play with the TacT and the new speakers.  Here is the latest measurement curve, after moving the speakers around a little.  Remember these are in-room measurements at the listening position, not nearfield or quasi-anechoic measurements like you see in manufacturer specs or magazine reviews.

I am running the main channels full range, as designed.  There are room suckouts at 54hz & 110hz.  Later this year the room will be expand ...

Why not reconfigure the impedance of the subwoofers so that you can easily drive them with any amp? They already are rated  9db higher in sensitivity than the mains and placing them towards the corners will increase the output another 6db.You've got plenty of sensitivity to burn.

[Rick Craig]

I'll go out on a limb and hazard the guess that you don't have suckouts at 54hz & 100hz but rather have gain just below those frequencies.  I.E. the valley is more the anechoic response tapering off slowly and there are four gain related humps from the room.  ~40hz, 80hz, 160hz, 240hz.  This is a guess but I'd bet a small sum on it.   If you were to bring these humps down with traps or whatever else then you sub towers would compensate for the slow slope from the main towers to give a flat sum.

Since the ports are on the front it would be easy to do a nearfield measurement on the subwoofers. If you place the microphone halfway between the center of the adjacent port and center of one of the woofers you can use the TacT to measure the subwoofer's anechoic response. This will allow you to see where the problems are with the room.

[JoshK]

Why not reconfigure the impedance of the subwoofers so that you can easily drive them with any amp? They already are rated  9db higher in sensitivity than the mains and placing them towards the corners will increase the output another 6db.You've got plenty of sensitivity to burn.

I thought that too.  Wiring them for lower impedance may draw more watts from your amp but it also lowers the dampening factor at the same time.

[ekovalsky]

Why not reconfigure the impedance of the subwoofers so that you can easily drive them with any amp? They already are rated  9db higher in sensitivity than the mains and placing them towards the corners will increase the output another 6db.You've got plenty of sensitivity to burn.

Rick, I thought of doing exactly that -- it would mean going from four 8 ohm drivers in parallel to two parallel groups of two 8 ohm drivers in series.  Nominal impedance would rise from 2 ohms to 8 ohms.  This would obviously improve damping factor and ease amplifier burden.  The former owner was contemplating doing this also but never did.  Sensitivity would obviously drop, but true efficiency would not change.

In room I am currently not getting an extra 9dB output from the sub towers compared to the mains' average SPL, but this may reflect the TacT's inability to deliver enough current into the very low impedance load.  Different class D designs deal with low impedance loads differently and I'm not quite sure how the TacT handles them.  It sounds like the Nuforce tolerates them quite well.

Although I have the bass towers near the corners, the drivers are still a good ways from the walls since the cabinets are 26" deep and the bases add another 8" to that.  Combined with their 7' height (plus 2" for the piano black bases and 2" more for the spikes, when I attach them) and the 13" width they are BIG.  And incredibly heavy -- after moving them, the VMPS RM/X which is reputed to be 375 lbs each, seemed light and nimble.

[JoshK]

The Tact will likely appreciate the 8R better.  Alternatively, if you didn't want to rewire (which is probably the better option, i'd think) you could try out the H2O amp.  It was designed primarily for running Apogee speakers so it is meant to handle low impedances.  It is a switching amp so the efficiency is high, which is good where you live.

[zybar]

Eric,

Why not try the Paul Speltz autoformers?

http://www.zeroimpedance.com/

This way you don't have to rewire.

George

[JoshK]

I'd be afraid that they'd saturate given the amount of current needed to drive such large bass towers.

[ekovalsky]

I'd be afraid that they'd saturate given the amount of current needed to drive such large bass towers.

That's why I haven't tried them.  Paul seemed to think they would work okay, but it just doesn't seem right putting a transformer with magnet wire in front of a subwoofer.  

I think the answer is a new amp.  There are plenty of AB and D models that can comfortably drive this load.  A local dealer is actually selling a Krell KSA250 which may be worth a look.  That would make a sweet bass amp though it does run hot.

[ekovalsky]

Since the ports are on the front it would be easy to do a nearfield measurement on the subwoofers. If you place the microphone halfway between the center of the adjacent port and center of one of the woofers you can use the TacT to measure the subwoofer's anechoic response. This will allow you to see where the problems are with the room.

I will do that and report back.

[ekovalsky]

It is probably low order plus a notch filter.

No notch filter according to Carl.  First breakup is at 5khz, which is over 10x the crossover frequency.

A crossover upgrade is available to take my Grand Ref II to III status (reported on by HP in the latest TAS).  I'm going to ask HP more detail about the changes he found.  Cost is only $1k so I'll probably end up doing this, unless I decide to go full DSP with quad-amping  

I must say it is nice that Carl has evolved the system from I > II > III status just with the crossover modifications (also new bases after the eary I models).   The crossover upgrades are fairly cheap and no changes have been made to the drivers or cabinets in the four years since this speaker's introduction (despite a corporate reshuffle).  This really helps preserve investment.  Vandersteen and Thiel are other standout companies in this area.  Owners of speakers whose manufacturers change drivers, cabinets, etc every year or two really take it in the pocket when they try to resell.

It is also great that no unsightly tweaks/tricks are needed to circumvent design issues to get maximum performance  

[doug s.]

I'd be afraid that they'd saturate given the amount of current needed to drive such large bass towers.

That's why I haven't tried them.  Paul seemed to think they would work okay, but it just doesn't seem right putting a transformer with magnet wire in front of a subwoofer.  

I think the answer is a new amp.  There are plenty of AB and D models that can comfortably drive this load.  A local dealer is actually selling a Krell KSA250 which may be worth a look.  That would make a sweet bass amp though it does run hot.

i would suggest checking out the power-wave series of amps from qsc.  they are relatively inexpensive (when priced from pro-audio shops), & make great sub amps, w/damping factors >500.  they also run wery cool...

http://www.qscaudio.com/

doug s.