Double stack ML subs?

[barjohn]

I am attaching my measurement of my system's bass response with ARC enabled of my Class 9's with 800X sub.

 

[ttocs]

I am attaching my measurement of my system's bass response with ARC enabled of my Class 9's with 800X sub.

So is this a single 800X sub with both 9's playing (L+R)? If so, that's pretty good! My room doesn't like a single sub, and when both fronts are playing (with no subs) a mono signal there's a deep dip about 10Hz wide right where you don't want it.

 

[barjohn]

Correct, a single 800X with both Classic 9's measured from listening position with REW.

 

[sleepysurf]

Finally had the chance to run ARC Genesis on the Anthem MRX 740, with my 13A's and stacked dual mono subs. Incredible! I finally have subs and mains perfectly time aligned, while eliminating the deep null at my main listening position! The default ARC setting starts crossing over the 13A's to the subs at ~100 Hz, so I want to try a lower crossover point. However, the bass is already tight and musical, without any sub localization that I can perceive. This was only my first run with ARC Genesis, so I haven't "tweaked" anything other than limiting DSP room correction to 500 Hz for the 13A panels.

I'll do a full measurement with my XTZ Room Analyzer II tomorrow, to check the actual speaker room frequency response, then start playing around with different crossover points. I also need to turn down the subs volume a tad as it's a little too "hot" in the lowest octave.

Thus far, the only downside to using the MRX as a preamp (replacing my beloved C-J CT5) is a slightly less "breathy" midrange (ostensibly due to the lack of tubes). I can live with that, as the significantly improved bass more than makes up for it.

 

[sleepysurf]

I ran XTZ Room Analyzer, as promised, but had some glitches with the measurements. I kept getting wonky results in the upper frequencies, as evident in this screen shot.

Not sure if it's a problem with the XTZ mic or a glitch with the MRX 740 (or ARC Genesis) processing. There's also a fair disparity between Anthem's "prediction" for the corrected curve, vs. what XTZ measured. Will troubleshoot further over next few days. Regardless, at least I've eliminated (most of) my former bass null, but still need to "tweak" things a bit.

 

[spkrdctr]

I'm late to the party but you guys did flesh out that you can't fill in nulls. You must move equipment/furniture etc. around to move or eliminate a null. You can easily pump a 1000 watts into a null to bring it up just a tiny bit. That is because the nulls are not perfect so you can move them but just a tiny bit with boatloads of power. So, it is not feasible. Very interesting thread. You can also see how even getting close to a good response curve radically changes the sound. Perfection is not required and probably not possible in a home setting. So, you guys take crappy rooms and tweak them using all the correct methods and the result is very, very good sound. Notice that it is NOT high dollar DACS, high dollar amps and wires etc. It is placement, angles, room correction, and room treatments. None of it is what is usually thought of by an audio consumer as important stuff. In audio, testing and working to fix issues far outweighs spending bug bucks on equipment. You guys are all proof...well some of you do have big buck equipment but that STILL needs all the same work to sound good. I'm impressed at the tenacity you guys show in going after great sound. You are all to be commended. What an enjoyable thread!

 

[JonFo]

I ran XTZ Room Analyzer, as promised, but had some glitches with the measurements. I kept getting wonky results in the upper frequencies

Alan, the anomalies in the 4 to 8K range are high-frequency room resonances as evidenced in the waterfall plot, where it shows retained energy (yellow 'mountains' extend further than at lower freq). This is the typical room ringing that occurs when a large line array is pumping out enough energy to engage resonances above 2K.

And it's not just the additive product of the rear wave, as even with that dampened, I've tested and found that the front wave alone (at least for the monster Monolith panels) will over-energize a room and cause ringing. It's one of the reasons I have so much absorption in my room.

The notch at 5.8KHz is likely a cancelation product from the comb-filtering caused by the rear wave, but could also simply be yet another room mode issue.

In your case, I might suggest re-running the ARC Genesis at full-spectrum and see if it knocks down that 4 to 8KHz range a bit.
But your best bet is physical treatments, your ears will thank you, but your wife might not

 

[sleepysurf]

Jonathan, I agree (somewhat) re the higher frequency room resonance and comb filtering. However, that never showed up (to this extent) with prior XTZ measurements using my C-J preamp with exact same 13A placement, so I'm thinking the Anthem's mandatory conversion of everything to digital (even if ARC is disabled) might be part of the issue.

Another factor is that mics are directional. Anthem advises their mic be directed up towards the ceiling, but XTZ advises theirs be directed at the speakers. As you suggested, I re-ran ARC Genesis (mic pointed up) with full correction to 20K, and then measured with XTZ using two mic orientations. With the XTZ mic pointed directly at the speakers, the highest frequencies were still "ragged." When oriented vertically, the measurement (see below) was smoother, albeit still rolling off. To be honest, though, I don't think my aging ears can hear that high anyways!

The blue trace (ARC disabled) illustrates the huge null I have at the main listening position. The green (with ARC enabled) shows how the null is filled in, with the stacked subs perfectly time aligned with the 13A's. Needless to say, it sounds superb from that standpoint. I'm still not entirely happy with the upper mids and highs, as they sound a bit "thin" and "tizzy" with ARC, despite measuring better. Some folks have commented (on other forums) that ARC can detract from overall musicality. I'll continue tweaking the ARC correction parameters to try and find a happy medium. At some point (before selling it) I'll swap my C-J CT5 back in to replace the Anthem, and have another listen. It's possible I might now be able to get acceptable bass (with subs stacked in corner) using the downloadable sub filters from ML, plus running ARC independently on just the subs and 13A's. More experiments to follow!

 

[JonFo]

so I'm thinking the Anthem's mandatory conversion of everything to digital (even if ARC is disabled) might be part of the issue.

I'd be very surprised, as A/D to D/A at those frequencies is a piece of cake, and I'm guessing Anthem uses good quality parts and implementation. Plus, the severity (and later correction) tend to point towards some acoustic issue amenable to electronic or physical mitigation.

When oriented vertically, the measurement (see below) was smoother, albeit still rolling of

More than the frequency response, the temporal behavior reflected in the waterfall is much smoother. Gone is the big dip at 5.8 and the decay seems shorter, meaning less ringing.
The roll-off depicted looks fine. A flat line to 20KHz is not pleasing to most.

I'm still not entirely happy with the upper mids and highs, as they sound a bit "thin" and "tizzy" with ARC, despite measuring better.

The 'Thin" part is due to less ringing, so less overall energy (thinner), but that excess energy is bad (slow decay).
The Tizzy part is hard to parse, as to me, that would mean added sound. Maybe the loss of the null at 5K (therefore more energy) is being sensed as brighter. 5K is in a range most describe as bright when boosted.

Some folks have commented (on other forums) that ARC can detract from overall musicality.

All DRC and even physical treatments change the sound rather significantly away from raw-room baselines. So yes, sounds very different, but losing the huge peaks and nulls is usually a win, and once acclimated, is more relaxing and natural sounding.

I disabled my DRC in the Media Room system last week to do some measurements and OMG, the sound was horrific with it off. Boomy bass, highs that rang, not musical at all.

That's not to say any and all DRC is good, as you note, tweaking is sometimes required. For example, adjusting the calculated delay to the sub output(s) can make a substantial positive difference.

 

[JonFo]

For example, adjusting the calculated delay to the sub output(s) can make a substantial positive difference.

Here are the examples of that taken last week.

First is a post-Audyssey Pro run, with the default delays for each of the two sub-outs.
Recall, this media room has 4 subs, 2 on the front wall, 2 on the back walls, fronts are on sub out 2, and rears are on Sub out 1 (s1).
Not as smooth as I've seen this setup before, so I knew something was off, and sure enough both delays are within a foot of each other (at around 11'), even though there is >9' physically between the front and rear pairs.
The range around 70Hz is over-boosted and 50Hz has a null.

So I changed the delay for the rear subs (s1) to be 7', and now it looks much better.
I tried a dozen different combos, and of that set, this was the most balanced. Not perfect, but sounds fine.

I probably overshot, as now 70Hz is low. Will try 7.8' delay on s1 next.
Point is: automated systems still need supervision and some critical review, and if necessary, manual adjustments.

Also note that the decay on both of these is exemplary, and that's from four cheap subs (Dynamo 300's) and minimal / cheap room treatments. In a room that's a near-cube (worst possible for low-end). Multiple subs at the right spots made all the difference.

And for those wondering, here is the baseline metric, with Audyssey OFF, yikes, not good. And that's with all four, a single sub looks even worse.

 

[sleepysurf]

Jonathan, that's quite impressive! Looks like +/- 5 dB from 20-120+ Hz. I presume those are REW measurements, using a Umik (or equivalent), taken after running Audyssey. Just curious, are those single measurements from the main listening position, or multiple mic positions averaged?

 

[JonFo]

I presume those are REW measurements, using a Umik (or equivalent), taken after running Audyssey. Just curious, are those single measurements from the main listening position, or multiple mic positions averaged?

Alan, yes, Post Audyssey (except for the last one).

Use a calibrated Niant mic held by the isolating holder and stand from my Audyssey Pro installer kit (that I use for the 8802 in the HT). So no mic resonances.

I use the latest REW for pre and post DRC runs.

Those are single measurements from the MLP. An average would look a lot better.

When aligning subs, I use Waterfall plots extensively, as the temporal behavior (how fast it decays) tells one a lot about phase relationships and modal behaviors. It is often preferable to optimize for balanced decay times, even if the Frequency Response looks a bit rougher.

The other plot I pay a lot of attention to when adjusting inter-speaker delays is the Group Delay metric in REW. Getting that to be as smooth as possible yields great synchronization. Important when aligning near-field vs far-field subs, such as the MBMs or the two rows of subs in the media room.

 

[sleepysurf]

I'd be very surprised, as A/D to D/A at those frequencies is a piece of cake, and I'm guessing Anthem uses good quality parts and implementation. Plus, the severity (and later correction) tend to point towards some acoustic issue amenable to electronic or physical mitigation.



More than the frequency response, the temporal behavior reflected in the waterfall is much smoother. Gone is the big dip at 5.8 and the decay seems shorter, meaning less ringing.
The roll-off depicted looks fine. A flat line to 20KHz is not pleasing to most.



The 'Thin" part is due to less ringing, so less overall energy (thinner), but that excess energy is bad (slow decay).
The Tizzy part is hard to parse, as to me, that would mean added sound. Maybe the loss of the null at 5K (therefore more energy) is being sensed as brighter. 5K is in a range most describe as bright when boosted.



All DRC and even physical treatments change the sound rather significantly away from raw-room baselines. So yes, sounds very different, but losing the huge peaks and nulls is usually a win, and once acclimated, is more relaxing and natural sounding.

I disabled my DRC in the Media Room system last week to do some measurements and OMG, the sound was horrific with it off. Boomy bass, highs that rang, not musical at all.

That's not to say any and all DRC is good, as you note, tweaking is sometimes required. For example, adjusting the calculated delay to the sub output(s) can make a substantial positive difference.

I'm learning the nuances of how MRX receivers apply ARC Genesis, so making good progress sorting things out. Turns out, by default, ARC applies room correction up to 5K. That's what caused the "tizzy" mids from my 13's. Running ARC Genesis in Pro Mode, I created multiple profiles, limiting correction to 200 Hz, 300 Hz, and 500 Hz respectively. All of them yield much more natural sounding mids from the panels, even if not perfectly smooth (as measured by XTZ), so I'll let my ears decide which cutoff sounds best. ARC Pro Mode offers additional customization of the target curves, including room gain, crossover points and slope, etc. etc., but I'm leery to "tweak" those settings without fully understanding what I'm doing! I need to do a lot more reading first. Will keep y'all posted!