SME

Thanks for your comments. I realize I will need to do the UHD upgrade sooner than later. It's a top priority when I have the income for it again. At the same time, Netflix by mail doesn't offer the discs, which is a drag being that the prices to buy the discs tend to be a lot higher too. This was an unusually quiet track, and so I'm surprised to learn that the Atmos track had an even higher crest factor. Wow! Even at "0" (or -2 vs. 85 dBC reference after dialnorm), the level seemed quite moderate on the DTS track. I assume I didn't exceed 128 dB WCS. Even if I did, I don't have that specific limitation in my signal chain. As such, I don't apply the static gains in the corrections and have no need to compensate using my MV. My headroom is effectively unlimited until the output stage to the amps (i.e., before the DAC), for which I have ~3 dB more than the signal that clips the amps with no load. On heavy infra content, the amps clip a fair bit lower than that, so in most cases, my headroom is closer to ~6 dB above the clipping point of the amps. This means that I can easily exceed 128 dB WCS with content above 20 Hz, but below there I run out of headroom a lot earlier. I've seen a fair few instances of heavy infra content clipping the amps, depending on my level choice and other parameters (HTTYD crash, TIH boxing / body slam?, a couple points in SW:TFA+BEQ), but this is the first time I've seen clipping on the digital output stage. That suggests to me that there were insane amounts of infrasonic content in those scenes, which doesn't exactly line up with what I see in the PvA. At the same time, you seem to indicate that the PvAs appeared very similar between the two tracks, apart from different overall levels. Is it possible that, even though the aggregate PvAs looked similar, the distribution of low frequency energy between the different channels was different between the two tracks? So maybe in the DTS track, the LFE was filtered a lot less than in the Atmos track, but the LCRS were filtered a lot more? If so, I can see how this could lead to a track that's much hotter in the infra after applying this BEQ. I guess I'll have to do my own analysis when I have the capability.

Sorry if I wasn't clear, but it's specifically *not* the bass extension that's the issue. The walls and ceiling provide plenty of boundary gain below a certain point. All my speakers are close to walls and have plenty of headroom below 150 Hz or so. Where the surrounds struggle is in the area of 200-900 Hz. The Volt 10LX at 95 dB is a very typical example of what's widely available. It's designed to offer bass extension that I don't need at the cost of sensitivity that I do need. Going from 92 to 95 dB sensitivity is not a significant upgrade. I'm aiming for the high 90s (or even 100 dB) in a 10" or 12" size, and I'm perfectly willing to accept significant LF roll-off and excursion limitations to get there. I'm curious what Erich comes up with, but the trend I see with most coaxials is that the larger ones just play lower and don't gain much if any sensitivity. Meanwhile, the vast majority of consumer speakers are just plain pathetic. As I develop my DSP optimization methods and learn as I go along, my opinions continue to develop in so far as what requirements for speakers are most important. It is my view that key among these requirements is high sensitivity/efficiency, especially in the low mid-range above roughly 150-200 Hz. Optimized DSP can fix a lot of speaker and room problems (even if most *current* products on the market don't do it well), but this solution requires EQ boost. Almost every speaker in every system has significant problems in the upper bass / low mid-range due to baffle effects and boundary interactions that require EQ boost to improve on, and anyone who likes a bass heavy sound needs more output in that range to properly transition to the subs or else quality suffers considerably.

I may have been overly paranoid, and I don't think I had the visual output indication (Motu 16A) that I do now. However, movies rarely have continuous, droning, high SPL sounds that are sustained for *a few minutes*, especially going to the surrounds. (The sub range is a different story.) Also, there's a big difference between playing very hot content for a few seconds vs. a few minutes. If you look at the average level for surrounds (not including sub bass) over one minute intervals across an entire soundtrack, I doubt you'll ever see anything higher than 90 dB per speaker, and I doubt you'll even see higher than 80 dB per speaker except in climatic scenes where the score is playing very loud. With that said, yeah I do want/need better surrounds soon bit more for instantaneous output capability than for long-term output capability. My current speakers are rated at 92 dB/2.83V/2pi/1m, but I'm fairly sure I'm clipping them from time to time with a ~300W/channel amp. (I can now watch the peak / RMS level of the signals going to them.) It's only on instantaneous peaks, and I don't hear any obvious distortion. Part of the problem is that they experience substantial boundary interference in the low mids, being that they are installed near a wall-ceiling corner. If I don't correct for this, they sound very thin and impart their thinness to the rest of the sound-stage when playing multichannel content including movie scores. With my latest optimization methods, I am able to correct this problem very precisely, but the correction does involve a lot of EQ boost. The end result is well worth it, despite potential for occasional clipping. For 99.9% of the time, the system sounds way better than it would without it. Still, I'd like to have a proper amount of headroom. FWIW, I'm pretty sure a lot of other systems have similar problems, and they most certainly apply to overhead speakers as well as surrounds. It has definitely affected my perspective on how to design multichannel systems. Surrounds and overheads are almost always placed in or near walls and/or ceilings. Where flush-mounting against a large, rigid surface is possible, there is less likelihood of a problem. However, for various reasons, this is often not possible. For example: It may not be possible or desirable to cut holes in the wall or ceiling. The room design may not allow a large rigid surface at the chosen placement location. It may not be possible to aim a speaker to achieve audience coverage goals when flush-mounted. For these and probably other reasons, surrounds and overheads are installed on or near the wall rather than in the wall perhaps in the vast majority of cases. With such placement, the sound of the speakers will be degraded without correction. The required correction requires a lot of EQ boost, usually to counter-act suck-outs in the low mids. At the same time, the nearby boundaries will usually interfere constructively in the bass range. As such when EQ/room correction DSP is intended to be used to achieve good performance, a good surround/ceiling speaker should have as high a sensitivity as possible. Excursion is not as important because most of the boost will be applied where excursion is low to begin with. Unfortunately, this limits the options considerably. Typical consumer surround speakers use small, medium sensitivity drivers and a ported enclosure to get the bass extension. These are seriously deficient in the crucial low-mid area. OTOH, small, high sensitivity drivers tend to lack too much in excursion and power handling. The implication is that good surrounds need bigger drivers, albeit light-weight pro-style drivers that have high sensitivity and modest excursion capability at the expense of bass extension. I had been planning on building new surrounds with 2 x 6.5" AE drivers and an SEOS horn, but I have other ideas now. In the long term, I want to build fully digital-driven arrays, but I have a lot of learning to do before I can do those. In the short term, I'm leaning toward a pro-style coaxial. I've seen some glowing recommendations for a few particular co-axials, for example from Radian. Unfortunately, most of the coaxials I've looked at, even the pro-style ones, appear to be designed to have significant bass extension I don't need at the cost of sensitivity. Most of the Radian coaxials just aren't as sensitive as I'd like, even at 12". IIRC, I saw one I liked from B&C that had a fairly shallow mounting depth and a woofer with a super strong Nd motor, providing like ~98-99 dB/1W in a 10". That's what I'm talking about! I think they also published polar response measurement data, which is almost unheard of in the industry. Once I have some expendable income again, I'll probably move on something.

In practice, how much does CMS actually vary in different manufacturing samples (assuming the same batch of parts)? I know we see lots of variation when we attempt to do measurements, but that's not necessarily the same thing. The value obtained may depend a lot on the details of the measurement, and the value is known to be quite sensitive to temperature and humidity. Both of these kinds of things should be expected to affect both drivers equally.

I don't see how a pair of drivers in series could cause a problem. Three or more drivers, yes, but only two? Anyway, I installed a 3/4" wall between chambers in my D.O. sealed sub because I was paranoid about local acoustic loading effects and wanted to option to possibly run separate signals to each end at some point in the future. That pic looks great! I'm sure your neighbors love you.

Do you know how similar the two tracks really are? I tested this out on the DTS track (no UHD here yet), and I'm seeing (post EQ) levels that are hotter than I'd expect, even for BEQ. One of the effects (SPOILER: select text to reveal <<< the destruction of Leia's ship >>>) clips not just the amps but my digital processor, which is pretty extreme. Note that my digital domain headroom is not based on 7.1 WCS but rather is a few dB higher than that required to drive the amp to max voltage. The PvA doesn't look *that* crazy. Are these just examples of extremely ULF-heavy broadband effects? Overall, the effects also seem a bit bottom heavy, and some of the ambiance and tension ULF seems rather overkill. Since I don't really how the tracks compare and can't measure them right now, I'm thinking of proportionally scaling back the correction until I have enough headroom and/or the bottom sounds a bit more balanced again. Perhaps it's another case, as with "Thor: Ragnorok", in which the two tracks are have substantially different bass profiles? Edit: I dialed back the BEQ to 75% (in terms of dB) and also made some minor broad shape adjustments (> 100 Hz), and the mid-bass slam is back and proper. I presume the latter changes were more important, but this BEQ was just too much for my system. Even at 75%, the aforementioned scene as well as (SPOILER: select text to reveal <<< Phasma falling into the flames >>> still clip in the amps at my chosen playback level. There's still plenty of infra power all over the place. Very nice! The overall level on the track is lower than typical, and I find it to be completely comfortable at MV "0" (equivalent to "-2" after dialnorm). The mixer must have liked less loudness or else his monitors sounded overly loud to him. Either way, the soundtrack quality is superb, easily better than TFA. Edit2: I watched this all the way through with the DTS 7.1 track at MV "0" (-2 with dialnorm) with guests and 75% of the correction applied. There was tremendous infra, but it honestly seemed a bit repetitive with almost every effect being very bottom heavy. The bigger on-screen events were rather disappointing because the effects were the same or weaker than for many other lesser events. Since I still don't know how close the PvA on the DTS track is to the Atmos track, I have no idea if my track was corrected as intended. Maybe the DTS track is ramped more aggressively or rolls off less in the ULF. The most impressive effect was the latter of the two scenes mentioned above that clips my amp, which unfortunately seemed only vaguely connected to anything obviously big on-screen. I don't know what frequency the effect focused on (mid teens?), but it was downright violent. I felt like my body was being tossed around like a jet flying through turbulence or something.

I recall "Ex Machina" had a loud droning score toward the end and the DTS:X mix sent the score almost entirely to my surrounds. I turned things down a bit because I was worried about long term power effects.

I got a conversion done. It's not really pro quality as I didn't do any dithering, but it should compensate for the mic cal down to 4 Hz: https://drive.google.com/file/d/1EylLR1mXkmaCI0mIjwQgPi7MkWVy0D1s/view?usp=sharing Anyway, the recording could benefit from some editing. It was hard to tell in a wave viewer where the good stuff is. It also has a couple of nasty pops, which I believe may be caused by EMI from the lightning flashes. I'm pretty novice with Audacity and didn't have any luck getting rid of them using its click removal plugin. Anyway, there's definitely thunder with very solid bottom in the recording. Oddly, the most impressive bass event doesn't seem to be caused by thunder. There's a lot of weird bass noise starting roughly around 56:30, and then a big ker-chunk right at 57:48. It's nearly -3 dBFS peak in the compensated version and almost entirely bottom end. Did you bump the mic or something?

Do you have a narrow band cal file for your mic? If you can send me that and the audio, I can create a compensated version of the recording.

LOL! If it's too loud, you're too old! It's like the equivalent of an SPL car in a listening room. But seriously, it's hard to argue with this point. If he actually ran the bass so hot as to use all the Othorn capability for movies, what's the point of having the Terraforms? Or are you trying to talk him into buying another 12 of them?

A 45% on Rotten Tomatoes isn't, like, "Conan: The Barbarian (remake)" bad. And it looks like it's not terribly long either. Something like 103 minutes of constant wind noise?

I see (now). Some details in the photo are hard to make out, what with all those huge boxes in the way. ? I think this is key. Models in Hornresp and whatnot are helpful to understand what happens in an idealized 2pi space, but when you stuff 8 Othorns into a tiny cinder block room whose dimensions are smaller than most of the wavelengths of interest, much of that goes out the window. How much footprint does that well and sump pump take up? I don't know if there is space for this, but what happens if you arrange them into two groups of four with all units firing into the front wall? Each group is two units wide and two units high and the width between the two groups is twice the width between each group and each side-wall. I'd also invert the unit on top so the mouths are closer to the ceiling, and maybe put them up on platforms to get them closer to the ceiling for better symmetry. Is there room enough in the left corner for that to work? I'm guessing the room is maybe 12 feet wide? So I guess that means 12" between the side wall and each group and 24" between the two stacks. Maybe that's too close. Or perhaps it works if there is enough distance to the front wall? Another option may be to cluster all eight (or just six?) subs in the center, still firing everything toward the front wall but forcing all sound to exit along the side-walls. In this configuration, the cabinets could be perhaps be angled to create expansions in order to reduce high order resonances and possibly boost output even more. Obviously, both of these options will work much better if the ceiling and right wall gaps are closed. Anyway, just throwing out more ideas here. I know if I had that kind of appetite for bass and a dedicated room, I'd probably just build my system into the structure itself.

EIGHT Othorns?!?! I don't doubt he broke windows. To be frank, if he is so concerned about getting the most performance possible from 8 Othorns, I suspect his health might legitimately be in danger, or else he just happens to like testing spaceship parts for launch worthiness in his spare time. To address the question directly, I am far from being an expert in these things, but what @StainlessSteve says here sounds right to me. The tapping effectively acts as a tune, where the horn between the driver and tap is 1/4 wavelength at the tuning frequency, below which native response drops with 24 dB/octave, plus the roll-off introduced by the requisite HPF to prevent unloading. In other words, any boost in extension achieved by reconfiguring the cabinets will be mostly overwhelmed by the natural roll-off of the systems. An optimized configuration could maybe help boost overall output though, and how the horns interact with the room boundaries is another variable. I think getting the mouths closer to the front wall is the right move, but I'd be reluctant to fire them into the center of the room because this will result in symmetric side-wall reflections that combine to create a deeper null along the center line of the room. Why not fire them into the side walls? This will minimize distance between the mouths and the most rigid boundaries. If smoother side-to-side positional response is desired, the horns could also be placed firing into the front wall in two-high stacks in a 4x1 configuration. Of course if the real goal is to get more extension, maybe he should just *upgrade* to Skhorn's which can be tuned a bit lower at the expense of some output. :)

By noise free, do you mean acoustic (i.e. fans), electric( i.e. hiss), or both? If you're open to plate amps, you might look at some of the lower power offerings from SpeakerPower. They work best with 4 ohm, however. I believe they are based on the ICE amps, so you may be able to find other amp designs based off of ICE in EU. (Most of the ICEs I think like 4 ohm more.) Finding anything that's acoustic noise free with that power level will be tough.

Can you run a slower sweep? That might help reduce the ripple. Otherwise, looks good. It looks like you hit the tune you wanted right on the mark.

The lower peak substantially depends on the suspension compliance, which is probably the least accurate T/S parameter given and the most likely to change with break-in as well as temperature and humidity.

Driver break-in may shift the impedance peaks ever so slightly, but it won't affect the tune, which depends only on the cabinet volume and port characteristics. Edit: I think you're closer to 21.5-22 Hz. The tuning frequency is precisely where the phase crosses zero.

I would add that the low-pass filter on the LFE channel will tend to soften any clipping that's present there anyway. The most offensive clipping is that which occurs on the mains channels, but I have found that the relative offensiveness of the clipping depends a lot on the quality of the playback system. If the playback system has sufficient headroom, a good balance between low and high frequencies, and no substantial mid or high frequency resonances, then most clipping is likely to be fairly inoffensive. My older playback systems likely suffered from all three of these deficiencies, and clipping in movies tended to be quite offensive to the ears. With my system now, and especially if reshape the response (where necessary) to compensate content created for X-curve calibrated systems, most clipping I hear just sounds kind of like a dirty recording, as though some of the signal isn't getting through. The sound quality is still obviously degraded, but it doesn't offend the ears like it did before. It also seems like most movies these days use soft limiting and/or compression instead of hard clipping like in the bad ole' days. That's good because most people have playback systems that are not so forgiving.

Is the YouTube version sanitized or something? Just going by ear here, but I'm not noticing much bass at all. There seems to be a tiny amount of content around 20 Hz, but otherwise it's just mid-bass and at fairly moderate level. It sure has a lot of distortion though. If that music is pushing your sub too hard, you don't want to even try playing a movie until you figure out what's wrong.

Thanks for the offer, but it'll be some time before I'm ready to do serious testing on other systems, at which point I'll need to be present to listen to the results myself. Of course, if I move to Europe or something, it could happen sooner. A flat in-room response will probably not give the best sound. At the very least, you'll probably want 5-6 dB of bass boost centered around 80-160 Hz. You may also want a more gradual slope (i.e. 1 dB/octave) a ways above and below that point. Feel free to experiment, of course. I'm not aware of any online samples from Floyd Toole's book, but if you are curious, I highly recommend buying it. In my opinion, it's the best book on the subject of sound reproduction, particularly for small room systems. A lot of stuff in there seemed counter-intuitive to me when I first read about it, but my experience over time as led me to change my mind and conclude it was correct.

Those look interesting. I'll definitely read them when I get the chance. The first article I opened (most popular) is dated 1998. Surprise surprise. This is hardly a "new" problem. The latest edition of Floyd Toole's Sound Reproduction book also has an entire chapter dedicated to discussing cinema sound where he remarks about the 2 kHz knee among other issues. Toole is consulting with an industry group that has recently been studying the problem with the aim of eventually establishing new SMPTE standards to address the problems with the X-curve. Toole's and Harman's recommendation seems to be to calibrate smoothed in-room response to a target that looks more like the Harman curve was is optimized (at least in theory) for "anechoic flat" speaker response cinema rooms and listening distances. While I think this approach would lead to an incremental improvement, it's still less than ideal. For starters, when people talk about frequency response, they don't bother to specify *how* the response was smoothed and *what form* was smoothed. These details actually matter a lot as far as the end result is concerned. For starters, one can choose a smoothing kernel which can be a flat-top ("moving average"), which is easiest to implement. Or it might use some kind of curve, for example a gaussian, which provides a smoother looking end result. As for what form, one can smooth the magnitude, the power, log-magnitude (i.e. the dB value), or the complex amplitude (where magnitude and phase are treated together as a composite 2D value). Power smoothing yields results most consistent with RTA and other continuous averaging instruments. Complex smoothing is actually mathematically equivalent to frequency-dependent window (!). FWIW from my testing, REW appears to do log-magnitude smoothing with a kernel that's similar to a gaussian. While it makes the data look prettier, it has a substantial effect on the end product. Furthermore, I have 99.9% confidence in the fact that fitting in-room response to any kind of target (even a room variable target) is suboptimal. Smooth in-room response should never occur in practice where any reflections are present, which means that in the process of EQing in-room response to be smooth, one actually introduces harmful resonances. How ironic! While I am a ways away from definitely solving this problem, I get the feeling I'm further along than anyone else at this point. It sure sounds like it.

Sadly, even a cinema that's "designed, implemented and maintained properly", seems to be a rare occurrence. While my memory may not be completely reliable, my impression is that cinema sound quality (at least here in the States) has trended worse over time. I recall being fairly impressed by new cinemas built in the 90s and early '00s, but newly built/renovated cinemas I've heard in the last few years since are meh at best. I believe a big part of the issue is the X-curve calibration standard and a shift toward stricter adherence to the target. The standard actually allows for a +/- 3 dB deviation, which gives a lot of leeway to a calibrator to allow more bass and treble and to smoother the knee at 2 kHz, all while remaining "compliant" with the standard. However, I suspect this kind of manual tweaking is not done much anymore as cinemas cut labor costs associated with setup and maintenance including calibration work. Many cinemas are probably calibrated entirely automatically. I'm not opposed to automation when it actually works, but for audio calibration, current technology is not up to the task. I'm trying to change the situation by working on better technology. I have a very novel approach that is showing enormous promise, but it'll probably be at least a year before I can test it outside my own room. It also remains to be seen how much tweaking it will need to work robustly in a variety of different rooms. I do have confidence that I could eventually adapt it to work in all kinds of rooms, which would be very exciting. Then the challenge would be to get the industry to actually update their standards, but in the meantime, it could probably be marketed to cinemas directly where it could provide substantial SQ improvements via download-able X-curve re-EQ meta-data.

You have a point. When I said "unfiltered", I left it vague. The context was ULF, which could be informally taken to be under 20 Hz. Extension under 20 Hz seems to be pretty rare these days, but there is a decent amount of stuff that reaches to 20 Hz or 25 Hz, as you pointed out. Of course studios have no obligation to deliver audio that's flat to DC (or 3 Hz?). Though it's nice when it happens. I wish they'd do it more often, for the tactile transducer users among other reasons. I know a number of cinemas have transducers now. Although they may be mainly of the "shaker" type with poor extension.

It's going to be a while before I do any more work here as I am *still* working on perfecting my system. My new calibration approach is much superior to the old, but I don't have a well-defined systematic approach yet. Different modifications lead to different different sounds, and I've listened to enough variations and developed enough intuitive awareness to know that I'm still not quite there. I'm getting very close though. It's getting better with each iteration. The detail I'm hearing in music is just crazy, to the point that many albums sound almost unfamiliar despite having listened to them on a variety of systems for years. I'm also leaning more toward waiting until I can make on-the-fly EQ adjustments before trying to do this. It's just very hard to evaluate changes with 30 second gaps in between and very time consuming to iterate repeatedly. I do want to make an interesting note. A couple weeks ago, I watched the BD re-release of "Monsters Inc". We started with 3D, which was fun until the PS3 choked (sadly, it has issues reading some discs), and because the PS3 doesn't deliver a TrueHD soundtrack and 3D together, I listened to the included 5.1 track. After switching to the other BD player (no 3D) to play the rest of the movie, I opted to continue the film with the 7.1 mix instead. The 7.1 mix was very different, and not just in terms of surround use. It was louder, and had quite a bit more mid-bass and treble (too much in both instances, IMO); whereas, the 5.1 had a bit of upper mid push. I'm fairly certain the 5.1 was a home mix with re-EQ. I'm not sure if the 7.1 was re-EQed from cinema or not. In any case, both my wife and I thought the 5.1 mix sounded better. The one benefit of the 7.1 mix was a bit more top end extension, but treble was definitely too hot. The 5.1 sounded more dynamic, had much better slam (by virtue of the mid-bass being better balanced with the low mids), and rendered ambiance much more faithfully. (FWIW, my experience is that the quality of ambiance rendering is a *major* clue as to overall tonal balance and sound quality!) Considering that ambiance is a *big* part of the appeal of surround, the 7.1 mix was quite a let down, IMO. By pulling back the upper mid just a hair, the 5.1 likely would have be nearly perfect. So if anyone wonders whether EQ is applied during home remixing, the answer is yes, at least sometimes. If anyone gets bored and wants to compare the 5.1 and 7.1 on "Monsters Inc", I'd be curious about others' impressions, but it's not a big deal.

You'll be surprised once in a while. Some TV shows do have some good ULF, albeit it can be spotty and inconsistent sometimes, and it's rarely as loud (relative to the dialog) as in movies.