Time alignment for Subs and Panels

There's a depression right around the crossover frequency (125 Hz) in both channels and in both Movie and Hotseat modes.  This depression seemed to respond to an adjustment (done in Movie mode) that was counterintuitive.  I delayed the subs LESS so that they sub gets the signal only 0.6 ms earlier (it was 1.6 ms earlier before).  That raised the output at the crossover frequency.

But this made no sense at all, since the sub is about two feet back from the panel.  Making the delay compensation much LESS than 2 ms should make it worse, not better, I figured.  But if I brought it down to 2ms, it got even worse than 1.6ms.  

Because of this anomaly, I decided to reopen this time alignment issue which I haven't investigated for about 6 years.

Back in 2020 I made 3 or more posts on the topic.  I tried various things, and concluded that you couldn't rely on measurements at the crossover frequency.  There were too many confounding factors, I decided, and measurements never made any sense.  I ended up concluding that simply using the distance between sub and panel was as good as anything you could do.  And that was the inspiration for the 1.6 ms delay, since the right sub (then) was about 1.6 feet back (back then).

Well that was before a lot of changes.  And I mean a lot.  I changed the speakers, I added the miniDSP FIR based linear crossovers, I re-did the EQ's many times (including just last week) and I moved the speakers farther out, about 2 feet from the subs now.  By the old reasoning, I should have increased the delay compensation to 2 ms, not reduced it to 0.6ms.

I decided this week the time alignment needed a whole new investigation.  And there's one thing I never thought of before.  The subs, which are SVS PB13 Ultras, (now, at least*) have internal DSP based processing.  All such processing introduces a time delay.   I asked Google and it replied I could expect a 5-6 ms latency for the DSP in an SVS PB sub.  That was based on an investigation by ASR (Audio Science Review) which I have a lot of respect for.

https://www.audiosciencereview.com/forum/index.php?threads/svs-subwoofers.17378/

(*When I first bought the SVS subs in 2009 they had all analog controls, and might not have used DSP, but I've replaced the plate amplifier boards since then, almost always under their warranty.  As of now, the right sub has the newest version I have, which is fully app controllable.  The left sub has digital controls, but I have to go to the back of the speaker to control it.  Since they are different plate amplifier boards, they might also have different latencies, though I'd expect them to be similar because SVS probably tries to keep the latency as low as possible since many people don't have a way of correcting it like I do.)

Since I never took the latency of the sub DSP into account before, it's no wonder the whole thing seemed confusing.  The expected delay compensation should be about 5-6 ms (for the DSP) plus 2 ms (for the 2 foot difference) or around 7-8ms total (which is close to my experimental result below).   

Furthermore, given the crossover frequency of 125 Hz, the response at 125 Hz should cycle through a range of values that repeats every 8 ms of applied delay compensation.  So if the correct delay compensation is 8 ms, it should also give a local peak response at 0 ms, which is close to the 0.6 ms I determined from a few quick and not very careful measurements I did last week.  (There is quite a bit of randomness in these measurements, and I didn't take them all the way to 0 ms because I knew that would be wrong.)

This does not mean that 0, 8, or 16 ms of delay would be equivalent in every way, only one of them would be the correct delay compensation value, but they would yield identical (or at least similar) response with continuous signals.

(I use pink noise, which is not really a continuous signal, but in some ways it acts like one.  Actually, given the fact that it is random, the very highest amplitude peak should be at the true correct compensation value, and the other local peaks at 8ms intervals should be increasingly lower, but they would be local peaks higher than at delay values in between.)

This seems to be confirmed by a series of measurements I did yesterday.  I played pink noise and measured the 1/6 octave band at 125 Hz while I changed the delay added by the sub EQ.  

(Note that I have the the panel delay fixed at 10ms, so if the sub delay is set to 2ms, the signal to the sub would therefore "lead" by 8ms because it has 8 ms less delay being applied.  By using a 10 ms fixed delay for the panels, which are equidistant from the listening position, I can compensate for the subs independently, since they are at slightly different distances, and I can compensate up to a 10 ms difference.)

I was starting at "9.42 ms" for the  the right sub which I had determined in more crude measurements last week.   (That would correspond to 0.58ms relative delay.). That made me guess that the correct setting would be 1.42ms (7.58 ms relative delay which is close to the predicted 8 ms).

So assuming my optimization last week was worth something, I started at 1.42 ms sub delay, tried 0.42, and then 2.52.  It was very confusing, and I ended up going all the way from 0.42 to 9.42 to see the pattern.  But the repeating pattern is unmistakeable.

Sadly my Analyzer app doesn't show the actual measured level after you freeze the analyzer.  (I let it run for about 30 seconds on random pink noise.). Instead, it shows the suggested EQ level.  As the response goes down, the EQ goes up.  So what we want is the lowest suggested EQ level, and since the bass level is generally raised for my room curve, this means we are looking for the largest NEGATIVE number (which corresponds to the highest actual output at the crossover frequency):

Sub Delay     Suggested EQ for 125 Hz (smaller means better--higher output)

9.42 ms           -1.0 dB

8.42                +0.6

7.42                +2.0

6.42                +3.5 ***worst***

5.42                +0.8

4.42                -0.6

3.42                 -3.0

2.42                -3.9   ***best***

1.42                -3.4

0.42                +0

This shows exactly the pattern I was expecting.  The amplitude goes through a sinusoidal cycle in about 8 ms (determined by the crossover frequency of 125 Hz, because 1000ms/125 is 8ms).  It reaches its best value (among the delays tested) at 2.42, which corresponds to 7.58 ms relative panel to sub delay compensation, right in the expected 7-8 ms range.  I would expect it to hit another peak at 10.42 (but in that case, the signal would be going to the panels first, which is clearly wrong).  The worst value is at 6.42, which is right in between 2.42 and 10.42.  At 6.42 ms delay the amplitude is so low it needs +3.5dB of EQ to reach the spectrum average, while most of the bass is above the spectrum average.  The output at the best delay tested, 2.42ms, is 7.4dB higher, as indicated by the suggested EQ of -3.9dB.

When I was making these measurements, I was getting very confused, not seeing the pattern correctly, so I mistakenly decided to a finer grid between 2.42 and 3.42.  I did not get a better result than at 2.42ms.  But it looks here like there could be an even higher value between 2.42 and 1.42, so I need to continue this investigation there to find the true optimal delay compensation value, or at least closer than just the nearest 1 ms.  My determination of a local peak around 9.42ms was probably mistaken, it would be more like 10.42ms.

(When I was making the measurements I often confused positive and negative EQ values, and I also figured that what might be important is to compare the EQ suggestion for 125 Hz with that at 112 Hz--that made things very confusing.  When I laid all the data out in a table it became clear that the expected pattern is shown most clearly in the actual 125 Hz value itself, which is where it should be clearest, because that's the crossover frequency.  It's still confounded by the fact that changing the delay does change other frequency bins too, which then affect the baseline the suggested EQ at 125 is computed against.  It would be nice if I could read the measured amplitude value rather than the suggested EQ value after the display is frozen, but I have seen no way to do that.)

While 2.42 ms is the best delay tested so far for the right channel, a quicker series of measurements suggested that 0.5 ms is the correct sub delay for the left channel:

Sub Delay            Suggested EQ

8.72 ms                -5.4 db

2                           -2.5

1                            -4.4

0.72                       -4.7

0.5                        -5

0.26                      -4.7

0                           -4.4 

Funny that my original 8.72ms setting is still better than 0.5, there may be a still better point between 0.26 and 0.72 but 0.5 is close, +/- 0.25ms.

The left sub is actually closer to its panel than the right sub.  Why it should need about 2 ms more relative delay difference (from the 10ms used on the panels) for time alignment is an open question.  I have three ideas:

1) The plate amplifier board in my left sub is a slightly older generation.  (The Left sub is a Sledge STA-1000D and the right sub is a Sledge STA-1200D.). The STA-1000D is controlled by a digital menu on the plate itself, whereas the later STA-1200D can only be controlled by an app.  However, online sources say they both use DSP and that all SVS DSP plate amplifiers have a 5-6 ms delay.  Nevertheless, it seems plausible that the later generation DSP has lower latency, since latency is something SVS would try to minimize.  If this is the correct answer, and my measurements are correct, the STA-1000D has about a 5ms delay, and the STA-1200D has about a 5ms delay, and I should go with my measurements in the expected range 0-3 ms (10-7 ms delay relative to panels).

2) Online sources could be wrong and perhaps the earlier STA-1000D may not use DSP under the settings I am using (I am only adjusting volume and setting the port mode).  In this case, I would be best using the 8.72ms setting for the left sub which only provides 1.28 ms relative delay, the delay being caused by having the sub about 1.5 feet back from the panel, with no digital latency.

3). I cannot measure the delay correctly with an amplitude measurement because the Left sub is in a corner while the Right sub is not.  Possibly the corner has a horn-like amplifying effect on an acoustically delayed response which overwhelms the direct response.  (This was what I concluded in 2020, not thinking about the possibility of DSP at all.). It is not obvious what I should do in this case.  Possibly adjusting for the maximum amplitude in the left as well as the right would be the best thing to do, in the 0-3 ms range.

I have emailed SVS customer support for info on the digital latency of the two different plate amplifiers.  If they cannot tell me I may try using one of the many SVS chat groups.

A test I could, in theory, do myself would be to change the crossover frequency to 60 Hz, and then compare the 8.72 ms and 0.5 ms settings.  If the plate amplifier has digital delay, this would ruin the response at 8.72 ms but leave it unchanged at 0.5 ms.  Because of the way I have the crossover patched in through FIR filters in miniDSP boxes, this would be a little complicated.  I would have to bypass or neutralize the miniDSP boxes and then add in 4th order LR filters in the Behringer DEQ's.  It would be complicated but do-able.

Direct measurements using digital recording would likely be more complicated.  I could create a series of 125ms pulses spaced at least 1 sec apart.  Then make a digital recording with high end microphone and see if I can figure out the time alignment.  It's difficult because low-passed subs only rise slowly from the beginning of signal onset, where as high-passed panels rise immediately then fade away.  Where the low passed sub signal begins is a judgement call, especially given background noise and non-anechoic conditions.  I've tried to do things like this before with very ambiguous results.

I've also previously tried to do time alignment using programs based on Maximum Length Sequences (MLS) like ARTA or the Tact 2.0.  That is complicated by the fact these programs were not designed to assess different speakers with highpass and lowpass crossovers.  My original MLS program,  LAUD 3.0, is now inoperable because it required 1998 vintage hardware like the Turtle Beach Fiji or Monterrey cards which have become inoperable due to capacitor aging, and even when they worked with a vintage PC, it took days and days to get them configured correctly on an ISA bus based system.  Even one of my tireless vintage PC friends (RIP) got tired of my endless attempts to get this working again circa 2016.  I once figured it took about 50 re-installs of Windows 95 to get one of these boards correctly configured, because the PC would "remember" the erroneous bus settings and you'd have to start all over from scratch to try all the possibilities and zero in on the one that works.  But I miss LAUD dearly because it gave much better looking impulse responses than ARTA. The MLS used by Tact 2.0 are basically simple pulses, LAUD and ARTA use quick sweeps which have more information, and therefore less noise, but ARTA fails to map this to a virtual square pulse the comprehensible way that LAUD did.  I've never been able to afford the true state of the art MLSSA system.  DIRAC, I don't even want to think about DIRAC.  I hate programs that try to do everything for you, including frequency response and phase correction, and just give you thousands of FIR coefficients instead of comprehensible information (typically loading them directly in your processor).  REW doesn't do time alignment type things any differently than an RTA like my iPhone analyzer app; it is meant for setting Programmable EQ's, not designing or time aligning multi-way speakers (at least this was true the last time I tried it in 2020).

Adding the deepest bass to the Movie EQ

 Bass was clearly rolled off in both channels below 31 Hz in the Movie EQ.  But I chose to leave it that way because I hate to use EQ boosts.  They can sound unnatural, they reduce dynamic range, and increase distortion.  Plus in some cases in my all digital system I can get digital clipping.

But when working on the Hotseat EQ this week,  I discovered I had previously been using boosts there up to 9dB (!!!).  In the newer version, now called Hotseat 2, I limited the boosts to 5dB because that worked well enough.  The subs are straining to get these levels at the Hotseat center of the room position, especially at 22.5 Hz, so I didn't want to push them any harder.

Digital clipping is rare in woofer EQ's because you never get 0dB signals below 125 Hz.  Such signals would destroy most woofers, so they are not allowed in recordings.  (I can tell you, having tried to release envelope pushing "Experimental Music" albums myself, streaming music distributors can get very picky about such things.  No clicks or pops are permitted.)

You could have near 0dB signals in the midrange, though I know for certain I reduced the digital level by -6dB in the panel crossover miniDSP's to prevent inter-sample-overs, which then means I could safely use up to 6dB boosts in the panel EQ's.  I'm not sure if I also did that in the bass EQ, where it's not needed to prevent inter-sample-overs, but the previous paragraph explains why there basically no issue with boosts in the bass EQ anyway.  I've never had an issue before, even listening to music at the 0dB level, the bass EQ meters never got close to clipping even with 9dB boosts.

And there's one more factor I'd completely forgotten about.  The Movie EQ has digital "gain compensation" set to -9dB.  Because the back of the room increases bass acoustically.  Meanwhile the Hotseat EQ (1 and 2) has the gain compensation at 0dB because you need all the bass you can get in the center of the room, where most primary room modes tend to cancel bass and you therefore have to pump out way more bass to get flat response there.  At least 9dB more.

This is not a trivial fact.  When I'm playing pink noise (recorded at -20dB) and playing it back at -7dB as I usually do for testing purposes, it sounds only moderately loud at the Hotseat position (around 87dBc or so) with a Hotseat EQ.  But if I walk towards the subs, they are playing very very loudly to the point I wonder if it is safe.  With the Movie EQ, they are just loafing along because it is so easy to get bass in the back of the room, and I hear no huge rumble when I walk towards the subs.

So if I use a +5dB boost in the Hotseat EQ, as I am now doing, that is equivalent to a +14dB boost in the Movie EQ.

So all this means that in the Movie EQ I can boost away in the bass EQ very freely, at least up to +14dB.

I found that +11.5dB boost in the left (sticking with 22.4 Hz and 1/3 octave bandwidth to avoid stimulating those pesky whole house modes below 20 Hz) gave me good near to flat response (with a slight rise) at 22.5 Hz, and +9dB worked best on the right.

But later I decided this was inappropriate for "Movie" mode, which I use when I have movie watching guests over,  mostly elderly people (like myself) including women.  They don't like to hear boomy bass either at the movies or in my home.  Many years ago a guest walked outside because the bass was too loud (that was when I was using something akin to the Hotseat EQ all the time, which did have terribly boomy bass in the back and in fact the entire periphery of the room).  The only quick solution was shutting off the subs altogether.  Later I was dialing them down by -7dB or so for movie watching, and then I'd often forgot to set them back when doing serious listening.  That were the events that motivated me to to have a separate Movie EQ which is automatically controlled by my Insteon/Universal Devices home control system.  (At bedtime, the Hotseat EQ is turned off automatically, and then to turn it back on I use either use either a mini remote or the switch on the Insteon module which controls it by powering a midi device which sends a midi signal to the Behringer DEQ 2496 EQ.  So when I start each day the Movie EQ is active, which is suitable for background music, and then to do serious listening at the Hotseat I press the button.  Sometimes I forget to do that but not for long, because I soon find I'm missing the bass.)

So I decided to take a more conservative approach for the Movie EQ, instead of a slight rise at 22.5 Hz, I'd make it a slight drop at 22.5 Hz (but not too slight either...since the Movie EQ is what I also use for background music).  Measured output is only about 3dB higher at 22.5Hz than it is at 1kHz.  Most of the bass measures higher and some bass measures lower (ideally all the bass below 200 Hz should be readjusted to be level with the 22.5 Hz measurement, which I think represents the best Movie "room curve" compromise).

Rolling back both boosts by 4dB seemed to give me the appropriate response, nearly flat but with a slight reduction at 22.5 Hz (and possibly 25 Hz as well).  The 22.4 Hz boosts are now +7.5dB on the left and +5dB on the right, all at bandwidths of 1/3 octave.

This is now the beginning of "Movie 2" where I intend to get flatter bass response all around using boosts or other measures.

Right Channel before deep bass boost

Right Channel with 9dB deep bass boost

Right Channel with 5dB deep bass boost

Left Channel before deep bass boost

Left Channel with 11.5dB deep bass boost

Left Channel with 7.5dB deep bass boost

I'm thinking now I could raise the Left Channel deep bass boost by another 0.5-1dB to make it match the curve in the right better.  And/Or I will have to see how well this works with actual movies.

Right now the biggest remaining issue in the bass seems to be the depression around the crossover frequency in the Right Channel.  I fixed a very similar depression in the Hotseat EQ by changing the time delay in the sub EQ.  I could do the same thing here most likely with similar results.  But now it bugs me whether that is the correct solution.  It makes no sense because the actual panel to sub distance now seems much too large for such a tiny delay.  And that was not needed in the Left channel which has a local peak at the crossover frequency.  Whether this is better fixed by delay or boost is now a reopened question in my mind.  I might try using larger delay difference instead of a smaller one too.  It seems to me that all the sub electronics ought to have more delay than the simple Class A/B amp used for the panels, and this would only add to the fact that the sub is further back.  Therefore, I would think that the subs would require much less delay than the panels, not just a tiny 0.6ms or so delay that fixed the right channel in the Hotseat position.

Balancing "Hotseat 2"

I wanted to check the subwoofer and system balances for the new Hotseat position, but ended up doing more.

I couldn't help myself but make a few more adjustments to the left sub EQ.  I deepened the cut at 144 Hz from -4dB to -5dB to remove a bump there.  I lightened the cut at 71 Hz from -8.5 to -7 dB to reduce the depressions in that range.  I lightened the cut at 39.4 Hz from -2.5dB to -1dB to reduce the depression there.

Then I made initial per channel measurements for the left and right.  I was wrong in thinking the right sub was lesser than the left sub, in fact the right sub was greater.  And it was noticable how I had boosted the left sub at 22.4 Hz and not the right.  So I boosted the right sub identically: +4dB at 22.4 Hz.  That gave both channels an identical 20-25 Hz "grunt" And setting the center boost frequency at 22.4 Hz, bandwidth 1/3 octave with amplitude of +4dB seems to evade some subsonic whole house resonances (which cause the glasses in the kitchen to rattle) which appear to be below 20 Hz.  But it makes the measured bass much more flat (the lowest bin in the iPhone "Analyzer" app is 22.5 Hz, so, yeah, you could say I'm adjusting to the app, while I don't like that idea in principle--it would be better to sweep with an oscillator as I have done before).

I was also bugged by the 280 Hz resonance in the right channel, which was visible both in the sub only and combined response.  To nix that, I added a maximum cut of -15dB at 280 Hz, where the right sub shouldn't be playing anything anyway.  That fixed the sub only response (of course) but only partly fixed the combined response, the room apparently amplifies 280 Hz in both.

I noted that the pink noise in the sub only had a very deep sound on the left, but limited in the right.  So I added a boost in the right sub identical to the 20.4 Hz, 1/3 octave, +4 in the left.  That made both subs sound much more similar, with an identical super low frequency grunt.

Despite all these changes, the right channel was +1dB over the left (the boost at 20.4 Hz made no difference at all, thanks to the C weighting).  So I went into the Tact and changed the L/R balance.  (The Tact digital preamp is the only place I can easily make that change, as the Behringer DEQ's frustratingly do not have individual channel gain controls, instead they have a stupid LR image shift which I almost never use.)

Previously I had cut the left channel by an astonishing 2.3dB.  I ultimately changed that to -1.2dB to bring both channels into exact balance, measured with pink noise and C weighting.  The subs alone are pretty chose in balance, with the left (!) measuring -0.2dB lower (not higher, as I had guessed).

To summarize, I changed many things in the right channel sub EQ, and in the Tact channel balance from -2.3dB on the left to -1.2dB.  The result is that both channels measure identical levels with pink noise and the subs nearly identical (0.2dB difference. favoring the right, when I had expected the reverse).

I saved these new "Hotseat" settings and now consider this to be "Hotseat 2".  (I can't keep calling this "New Hotseat" as there will surely be many new adjustments.)

There remains a slight audible difference between left and right.  The right seems to have more mid highs, while the left may have more high highs.  This could have to do with the adjustment of the "HF Balance" controls.  I can easily access the HF Balance control on the left channel (it has a knob in the interface) but I cannot easily access the HF Balance control on the right (it requires moving a bracket inside in interface since it is the later design).  I don't think the difference between interface types is very important because the older one (left channel) was give the "C" mod by a famous Acoustat modifier, and "C" mod was the factory version on the right.  The mod level is what makes the most difference in frequency response.  The fact that the right channel has the "Medallion" transformer and the left channel does not is something I consider to be unimportant in frequency response measurements (the Medallions have better overload characteristics, which is relatively unimportant as I'm crossing over at 125 Hz which keeps the bass transformer from overloading as it often does if you are running the Acoustats full range).

Or it could be that the right channel just fires into a different acoustical space than the left, explaining the difference.

It is incredible how many changes needed to be made since the last "Hotseat" adjustments which were done just last year.  I have only changed the acoustical space slightly because the old recliner chair is no to the immediate right of the listening chair, which is now the Director's Chair.  Perhaps the ceiling Q has changed because of foundation shifting and roof replacement.  I've changed the Acoustat positions only very slightly and the subs none at all.  None of these ideas seems like an adequate explanation.  Perhaps I should run an RTA at least monthly or at least quaterly to be sure the system EQ is still working well.  The new left channel response is pretty smooth, good enough for me to say (as we used to say at Audio Dimension when I worked there) "Ship It!"

Right channel response 10/5/26 Hotseat 2

Left Channel EQ Hotseat 2

Right Channel EQ Hotseat 2

Afterwards I listened to parts of tracks on "Bass Ecstasy" by Bass Erotica, and "We Want to Be Loved" by Grouse.  The latter sometimes gave me goosebumps and brought tears to my eyes.  THAT's what high end audio is all about.

Adjusting the New Hotseat EQ's

To make more precise measurements, I set up the microphone stand with in the new hotseat position, with the director's chair in place.  The adhoc measurements in October were using the iPhone 16, whereas I routinely use my iPhone 8 on the stand (so I can keep it there).  Comparing iPhone 8 and iPhone 16 on the stand there are no substantial differences (whereas the previous iPhone 16 measurements in October while being hand held are a bit more different).

Left Hotseat, stand, iPhone 8

Left Hotseat, stand, iPhone 16

Left Hotseat, handheld, iPhone 16

When the iPhone was handheld (that measurement was done on October 19) there is somewhat more high frequency loss in the 2k-15k region (which seems reasonable) but about the same (or somewhat less) at 20kHz.

The high frequency loss generally seems to be the result of wide angle with the face of the speakers.  Since they are "aimed" to be off axis slightly in the back of the room there, they are even more off axis in the front.  I can tell as I walk backwards from the hotseat position to the back of the room that the highs are increasing.  The system measures much flatter to 17kHz in the back.

I can't change the angle as it's a compromise between hotseat and back.  Aiming the speakers more towards the hotseat position destroys the imaging in back.  As it is now, the high frequency loss at the hotseat is not objectionable, it might even be musically beneficial.  Voices and acoustical instruments sound "real" and not electronic as they often do with hifi systems.  I also note that it doesn't entirely reflect what the ears "hear" because the ears are closer to being on axis (the microphone is positioned to the center of the head, and not where the ears are), plus the ears are highly sensitive to the sound coming in perpendicular to the ears, which is not as much affected by the angle with the speakers.  So it "sounds" higher from 2-10kHz than it measures.  If I made it flat it would no doubt sound too bright.

I'm not sure this problem warrants "fixing" now by EQ or other means.  If I use EQ, I'd have to start switching panel EQ's between hotseat and back, an additional complication.

The somewhat lumpy bass probably requires a re-do rather than a tweak.  To do it correctly I need to get out the oscillator again and do sweeping, with a remote EQ at chairside to fine tune the PEQ centers and Q to get them just right.  The RTA can be misleading for such fine adjustments.

Unlike in the back of the room, there is no suckout in the midbass between 125 and 250 Hz.  So the panels do not need more boosting there (or, conversely, less sub cuts, I seem to have added rather large subwoofer cuts above the crossover frequency.  These cuts  appear basically fine for the hotseat even if they may be overdone for the back. 

I was going to call it a day, then I noticed something bad.  Right where there is that awful triangular peak at 90 Hz, there is also an apparently outdated boost!  I removed that boost and it rather much smoothed the peak down.

So then, to help fix the depression at 100 Hz, I moved the cut at 108Hz up to 112 Hz, narrowed it slightly, and reduced.

To help with the depression at 56 Hz, I removed the cut at 49 Hz.

The bulge around 35 Hz was explained by a boost there.  I removed that boost, and added a new boost at 20.7 Hz at as much as I dared (+5dB) even though higher measured better.

I added a new boost at 60 Hz to fix the depression on both sides of that.  As much as I dared (+5dB).

Each change resulted in better looking response.  Now it looks more like I am aiming for in the bass, relatively smooth from 250 Hz on down with a slight rise.

I saved the new settings in the left sub EQ to the Hotseat EQ.

New Left Hotseat sub EQ's

I worry mostly about the two new boosts.  The lower one might stimulate subsonics (which I can't measure with iPhone) too much.  I should sweep to be sure stuff doesn't go badly wrong.  But this new EQ looks good.

I wasn't going to make any adjustments today, but after I found the obvious one I just had to keep rolling.

Playing Bass Ecstasy by Bass Erotica and writing this up in the kitchen it was clear there was too much wall rattling.  I fixed that changing the bottom boost to 1/3 octave instead of 1/2, and then raised it to 22.4 for good measure.  That should also help keep tonearm resonances out also.

Two days later, I rolled back the 5dB boosts to 4dB boosts, reasoning that it makes very little difference in the response, and the lesser boost significantly improves dynamic range and lowers distortion.  The new response looks like this:

Left New Hotseat 11 04 26

Left New Hotseat EQ 11 04 26

A few hours later I'm thinking I'd like to both increase and retune the boost around 60 because the subs have plenty of dynamic range in that area and it's notably depressed at 56 Hz, though still much smoother than before.  I'd also like to fix the depression around 100 Hz, but I already eliminated the cut there and narrowed and raised the cut now at 112 Hz.  But it's way better than when I started, and that kind of fine tuning should probably include some oscillator sweeping.

Just after those minor changes, I decided to tackle the Right Sub EQ in a similar way.  It had a huge peak around 35 Hz (where there was a boost in the EQ) and a huge depression around 45 Hz (where there were two cuts in the EQ).

Right New Hotseat Before Changes

Removing the 35 Hz boost made it a lot smoother.  But even removing both cuts around 45 Hz failed to make that depression much better.  The subs have plenty of dynamic range around 45 Hz so I just added a +6dB boost right there, not entirely fixing the problem but helping a lot.

Right New Hotseat 35 Hz boost removed and 45 Hz boost added

Now the biggest problem seemed to be the depression at 125 Hz.  Removing the -2dB cut at 127 Hz hardly changed anything.  However I realized the crossover is around 125 Hz (or is it 120 Hz, I can never remember) and I wondered if changing the sub delay would help.  The subs were delayed about -1.6 ms from the panels.  I tried decreasing the delay and found a sweet spot around -0.6 ms.  With the LR crossover, both drivers should be in phase at the crossover frequency, so maximizing that point out to be the correct delay, so I stuck with that.  The distance seems to be somewhat greater than would be suggested by 0.6 ms, but possibly the EQs tend to mess up the delay a bit, so I'm sticking with the 0.6 ms for now.

It was still a bit depressed at 160 Hz, so I removed the sub cut at 152 Hz, making the overall response much smoother than before, so I decided to call it a day right there.  Way smoother than before, though I believe it could be made still better.

Right New Hotseat after > 125 Hz changes

Right New Hotseat EQ 11 4 26

Right New Hotseat Delay 11 4 26

Note that the delay for the panels is "standardized" at 10 ms, so a delay in the subs of 9.42 means 0.58 "less" delay.  Prior to this adjustment it was 8.42 and 1.58 ms "less" delay.

Things I need to do next include checking the sub balance.  I was going to do that but ran out of time and energy for the day.  The subs also need to be re-swept with an oscillator, and it would be good to do another sweep with both subs and panels.

I'm thinking the Right Sub level could be raised a few dB, raising the depression around 45 Hz

New Hotseat Position Measurements

 I'm calling this the "new Hotseat" because it's now a lowback chair (which makes for better imaging and measurements) and moved slightly farther back than before (because the speakers are now angled slightly more outwards than before, for better imaging in the back of the room).  

The panel EQ for the hotseat position is currently identical to the one now used in the back of the room (I haven't yet wired up separate EQ's for panel frequencies...which could be done but I haven't yet seen the need).  The EQ differs only for the "subs" where it is extremely important.  Flattening the bass at the center makes it way too strong in the back of the room, and flattening the bass at the back of the room makes it far too weak for the center of the room, generally speaking, so there need to be different bass EQ's--no question about it--if both positions are taken al all seriously.  The back of the room EQ also works pretty good in the sides of the room, where people watch movies so I call it the "Movie EQ"

The measurements don't look quite as good as when I did them two years ago.  (Note: these are just ad hoc measurements without a stand so far.  When I need to get really precise, to show the effects of small changes, I will start using a stand again, which takes considerably more effort.)

Left Channel New Hotseat

Right Channel New Hotseat

The left channel has a weird triangular shaped peak around 90 Hz where the right channel only has a modest "rise" (which in context looks more like depressions on both sides).  But (unlike the back) the lowest frequency bin (22.4 Hz) is not depressed compared to the midrange, only slightly depressed from the region around 40 Hz (where a huge modal peak is being suppressed by EQ).  Somewhat troubling is the rolloff above 1k which shows the -3dB engineered "Peterson Dip" (inspired by the Gundry/Linkwitz Dip) but only on top of much deeper lowpass shelf.  I think the lowpass shelf is largely resulting from being off axis from the speakers, which is hard to fix without messing up the imaging in the back of the room.  It doesn't sound as rolled off as it looks, there is a wonderful 3D image but only behind the speakers, sounding just slightly "distant" and never forward or aggressive as the Acoustats can sound on axis without EQ.

The original and still used Hotseat EQ has greatly increased bass compared with the Movie EQ, achieved through both higher "Gain Compensation" level and a few actual boosts (up to 6dB).  I don't clearly remember doing so, but I must have allowed for up to 6dB boosting by having a 6dB cut in the MiniDSP's which do the phase linear 8th order Linkwitz / Riley crossover.  It was absolutely necessary to do that in the high pass panel DSP because I could get +6dB ASRC* inter-sample-over peaks and clipping otherwise, but it not in principle necessary to do that in the lowpass bass DSP because low passed bass doesn't stimulate ASRC clipping. 

 (*Digital is not as simple as you think especially when there is Asynchronous Rate Conversion (ASRC).  I would have greatly preferred fully synchronous digital transmission in which the data is unchanged, but the miniDSPs don't have that option like my older Tact preamp.  But I've turned the defect into a feature by reducing the miniDSP digital input levels by 6dB which then permits up to 6dB boosts in the downstream DEQ's without risk of clipping.  Inter-sample-overs can occur when new datapoints are interpolated in between existing near-clipping high frequency datapoints....this almost never occurs in audio signals except with test signals designed to elicit the effect like those used in jitter tests--where there are often maximum level waves at the half the sampling rate.)

Hotseat EQ gain levels

Movie EQ gain levels

Hotseat Right EQ

Hotseat Left EQ

Not only am I using boosts in the Hotseat sub EQ's, I'm running the EQ at 0dB whereas I'm running the Movie EQ sub EQ at -9dB.  This is because generally speaking the back of the room needs a lot less bass.  However, that also means that I could boost the bass where needed up to 9dB or more in the movie EQ, which might be useful at the very lowest frequencies.  I'm already boosting the bass by up to 6dB in the Hotseat EQ, which makes it effectively up to 15dB higher.  Not knowing this, I had been shying away from using boosting in the Movie EQ even when I thought it might be helpful.

Spooky Good Imaging in hotseat position with lowback chair

For a couple of years now I've been optimizing the living room stereo for listening at at the center back of the room.  I created a new EQ for this called the "Movie EQ."  By necessity it had to be very different from the "Hotseat EQ" for listening closer to the speakers.  The Hotseat EQ sounds very boomy in the back of the room.  The periphery of the room magnifies the room modes, whereas the center position tend to cancel them out too much.

During this time I haven't bothered to sit at the old hotseat position.  I've loathed the idea of dragging my full back recliner back into the center position, especially since I try to show at least one movie a week.  I also feared all my optimizing for the back of the room had messed up the sound at the old hotseat, though I retained the original Hotseat EQ for the subs as a separate memory (which can be quickly loaded at pushbutton through my Insteon/Universal Devices home control system).  But I have no separate EQ for the midrange which has gone through a lot of changes too, and changes I made to the speaker angling cannot be easily reversed.

I bought a new lightweight low back folding chair for the hotseat position, but hadn't gotten around to trying it yet until the weekend just past.  I feared the hotseat position would be all messed up now.  But I started trying it on Saturday and then on Sunday found a truly wonderful position, about further 6 inches back from the original hotseat position with the recliner.  The new position is probably required because I changed the speaker angling to solidify the center image when listening in the back.  The speakers are less canted inwards, which is necessary to get a stable center image at all in the back of the room.

Listening to Citizen Steely Dan Disc 3, in this new hotseat position with a lowback chair, with the hotseat EQ selected, the imaging was Spooky Good.  It seemed like Donald Fagan and all the other musicians were right there in front of me. Nothing I've heard elsewhere for a long time has given me that same feeling.

The back of the room "Movie" listening position simply does not and cannot compare with this.  A much wider angle between the speakers is required, and there may be other advantages of being close up to the speakers.  There is much more direct sound from the speakers.  Some have described this as being like "Giant Electrostatic Headphones," even when both speakers are still in front.  (There are problems with trying to put both speakers exactly on the sides for stereo, though it can be interesting to play with.)

I do really need to use the hotseat position and EQ for serious listening, which I should do more often.

At the new hotseat, the speakers are about 45 degrees to my left and right on either side.  I'm about 5 feet from the plane of the speakers, and they are 5 feet apart.  But where to measure this is tricky.  If I measured from the middle of the speakers they would be over 6 feet apart, making the angle greater than 45 degrees.  But I now believe what is most important is the angle to the closest part of the active membrane, which is only about 2 inches inwards from the outer edge of the speaker towers.  That angle is somewhat less than, but still close to 45 degrees.  I would guess it is about 40 degrees.

I think the relative speaker position and angling is the number one most important thing in getting Spooky Good Imaging.

It was my friend Tim who originally told me, over 15 years ago, that I should experiment with sitting close to the speakers at a very wide angle, greater than the 30 degrees often suggested.  He was right.

He also said the exact position of the speakers and angling might not be the same for every person, because our outer ears are not the same.

If you keep the speakers at a fixed position, you move forwards along the center image to find the place where the center image disappears.  That means the angle is too wide and you are too close to the speakers.  So then you move backwards until the center image reappears.  Around where the center image reappears is the best possible position (or angle between the speakers).  Further back, and you get proportionally less direct sound, and the image becomes more diffuse.

That was essentially what I did on Saturday to find the new listening position.

Alternatively, if the listening position is fixed, you could move the speakers in and out from the wall, but for me and my speakers that would be exceedingly challenging for both physical and technical reasons.  I have the sub to panel delays dialed in precisely, for one thing, and changing the panel positions would mess that up.

Even if you ultimately end up moving the speakers, it's far easier to experiment by moving yourself closer and farther from them.

In the new position, the center is so perfect and well focused it's clear that my optimizing the speaker position, angling, and delays for the back has not messed them up for the hotseat.  If anything, it's far better than it was before now.

I purposefully did not angle the speakers outwards as much for the back as might have been beneficial there, hoping it would not mess them up for the hotseat, and my compromise for the back has indeed worked out perfectly for the front while being "good enough" for the back.

*****

I think having a lowback chair is good too.  When you are sitting in a high back chair, you get near reflections from the back of the chair in your ears.  These reflections matter more for imaging than for frequency response, but the frequency response is affected too, as I noticed when doing measurements.  I generally have always done my measurements with the chair in place because that is that is what you do when listening.

There's a long story here.  When I bought the current chair around 2008 I intended to take it to a furniture repair place and see if I could have the back shortened.  I knew this would be expensive but I figured it would be worth it.  I never got around to doing that.

The cheap folding chair I bought, which has flexible and ventilated plastic seat and back, is adequate but not as comfortable as I would like.  I think I will explore other options, like a director's style chair I already have, or others.

*****

The Citizen Steely Dan series became available in 1993.  While many people think of it as merely a repackaging of Steely Dan albums, it's a whole new mastering that was never heard before.  You knew about this if, like me, you read Steely Dan blogs around that time.  I also knew something like this was needed for a long time.  I don't know how it compares with later Steely Dan CD's, I suspect any high resolution versions were made like this too.

It turns out that one of my best friends was a techie hanger with Steely Dan.  He once invited me to the back room for a Steely Dan recording session.  I passed on this opportunity around 1976.  I had never heard of Steely Dan before.  Oops!

Later, I missed a chance to dub a copy of the master tape Steely Dan was circulating among its members.  I could not get there on time.  I ended up getting a second generation cassette tape dub (using my Nakamichi 600, and copying a tape another friend of mine had made on his Nakamichi 600) and it was a revelation.  A whole world of missing detail was on that tape that I never heard from the vinyl records.

(I think I can hear my friends voice on the master tape, and possibly on at least one song on Katy Lied.  You can also hear his girlfriend, who had originally been Denny Dias's girlfriend, playing some auxiliary percussion.  I could have been there too, dammit.)

For quite awhile I felt this was because tape is superior to vinyl.  But that wasn't the reason,  The reason was that the circulating master tape lacked the final Dolby processing and reverb that was added to the vinyl pressing (and also to early CD's).

There was a big technical problem here.  Steely Dan did all their recordings from at least Katy Lied onwards with dBX.  They needed dBX because of the incredible amount of overdubbing they did to make the final albums.  But the record companies only used Dolby A when making the final masters to be pressed on vinyl records.  So the final albums were retouched with EQ, reverb, and other tricks to sound OK and cover up the error in decoding.

In the Citizen Steely Dan recordings, they went back to the original masters, made with dBX, and decoded them with dBX the way they should have been.  Then they added the touch of reverb that may be less heavy handed than in the original vinyl.  The result is Steely Dan the way it should have been.  And it is incredible.  Even better than my second generation tape dubs, which have a lot of tape hiss (my friend and I chose not to use Dolby B at all).

This may be more relevant to some of the discs in the Citizen Steely Dan series than others, but it appears to be especially true for Citizen Steely Dan Disc 3.  I'm not sure if all the albums had the processing error.  I think it was true for the albums from Katy Lied to Aja.  I have never heard the songs on Disk 3 sound as good as they did last Sunday night.  It was Spooky.

*****

Among the changes I made when tuning my system for the back of the room listening position was changing the relative channel levels and delays.  I didn't know how this would work for the hotseat position, but I now know it worked out great for that too.

*****

I worked for 3 months on the subwoofer EQ for the hotseat position in 2024.  I ended up making giving it very smooth and nearly flat bass response.  Probably for many reasons, it is not as smooth now and could use a bit of rework for the new hotseat position.  But still, it is incredibly good.  It gives a somewhat higher bass level than I achieved for the Movie EQ in the back.

I wondered how this could be, given that the Hotseat EQ has many more and deeper cuts to eliminate room modes.  Well, there are two reasons.  The gain compensation for the sub Movie EQ has a -9dB setting for gain compensation.  For the Hotseat EQ this is 0dB.  I also added a 6dB boost for one frequency range in the Hotseat EQ.  Neither of these changes could simply be applied to the Movie EQ, but I might try doing similar things.

Laser Measure

 I read about audiophiles using laser measuring devices on Stereophile and even say they're essential for serious audiophiles (to get your speakers precisely positioned) I've never bought one.

This looks like one I'd like to get, a Mileseey S7.  

https://mileseeytools.com/products/outdoor-laser-measure-tool-mileseey-s7?currency=USD&sku=18063677008810900264450910&utm_medium=cpc&utm_source=google&utm_campaign=Google%20Shopping&utm_source=google&utm_medium=PMAX&utm_campaign=S50&gad_source=1&gad_campaignid=23050170196&gbraid=0AAAAAoVRmp6WhX8rcXFFIy1Ypabp95nu0&gclid=CjwKCAjwqonVBhA4EiwA9wYJ3f5cErJTQ0TCJu2b96xTEdEz3BYFNRhfXWsYMBbCU4E1EmUYO8xjYxoC2y4QAvD_BwE

But since my income is limited, I plan to wait until I've sold surplus audio equipment of equal or greater value, which I really need to do right now for other reasons too but have been too busy with urgent household projects recently, mainly that I want to be prepared for the rain storms likely with the Super El Nino and have been working on critical hardscaping and drainage features all around the house, stuff I should have thought of 30 years ago before my house needed foundation repair, now I want to prevent it from needing even more foundation repair, fence repair, driveway repair, etc.