Sunday, February 27, 2011

Living Room Dimensions and Room Modes

Had friends over for my monthly discussion party tonight, so I was able to make measurements of living room (which requires second person on opposite wall, since there are no straight paths on the floor).

Length: 187.5 inches (15.63 ft)
Width: 157 inches (13.08 ft)
Height: 98 inches (8ft, back), 96 inches (7'10" front), 110.5 (9.21ft, center)

I'm running the program ModeCalc from RealTraps, which produced the plot above.  I used 8.6ft average height.

Length has modes which are multiples of 36.15 Hz (36, 72, 108, 145, 181, 217, 253)
Width has modes which are multiples of 43.20 Hz (43, 86, 129, 172, 215, 259)
Heigth has modes which are multiples of 65.70 Hz (66, 131, 197, 263)

Total volume: 1758 cu ft
Ratios: 1: 1.52 : 1.90

Note that the height is an average one, so the 65 Hz sequence, and the ratios, are fuzzy.

If I chose to use max height, for example, the ratios would be:

and the height modes would start at 61.35 Hz.

Given that all rooms have to have modes, it doesn't look bad, the modes are reasonably well spaced and don't tend to pile up much.  You notice in all such plots that as you get to higher frequencies, you get more mutual node re-inforcement because on a log frequency scale the frequencies get closer together.  So there's a gradual increase up to the maximum calculated 500Hz.  Below 125Hz, there is only one area where two node-multiples come together.  That is the 2nd harmonic of 36hz (72 Hz) and the primary height mode of 66 Hz, and the height mode should be somewhat soft because of the center-peaked ceiling.  In fact, I have noticed a mode in that area.  Though RoomCalc doesn't show it as a pileup, you could also argue there is a pile up because of the two lowest principal modes at 36Hz and 43hz (caused by room length and width).  That's actually a 1.2 ratio, which is about the smallest acceptible such ratio.  Then there is a more pronounced pile-up at 130Hz (2nd harmonic of room width by third harmonic of height).  Then the next  next pile up is with 215 and 217 Hz.  Even "good" ratios have a few pile-ups like that.  But in my system, dipolar speakers play above 85 Hz and their output seems to stimulate room modes less.

So, not that I could change the dimensions anyway, they don't look too bad.  It looks good enough that you might believe the house developer (low cost San Antonio builder Rayco, which sold out to national developer K&B about 10 years ago) actually considered the acoustic properties in the floorplan, which has no doubt been used countless times.  No, it's not one of the "optimal" ratio sets, but it's not bad either.

Anyway, I now agree with what Real Traps says.  Regardless of how the room physical or acoustic measurements work out, if you are starting from a pre-existing room, about all you can do is add as much bass trapping as you possibly can.  So that is one of the next projects.  It will also be useful to run before-and after measurements to see the affect of different materials and different placements.

Thursday, February 24, 2011

So many traps, so little time...

[This post is under construction.  I am trying to consolidate and summarize the difference between competing bass traps here, but it takes time to gather all the information.]

I have become convinced of the need for bass trapping, and basically agree with the assertion by Real Trap's founder Ethan Winer that it's almost impossible to have too much bass trapping (and this is much different than HF absorption, where you can definitely have too much and maybe don't even need any).

Maybe the walls themselves should be designed as bass traps.  Thicker, with solid masonry outer wall, stuffed with fiberglass and lossy plastic membrane surface instead of gypsum board.

I'd love to get the high performance Megatraps from Real Traps, but they are monsterously big and I can's see right now even where I could put them.  (The price is very reasonable for the performance provided.  I just meant I don't want to buy something I might not ultimately be able to use because it's so big.)

Here are the comparison measurements of various Real Traps units compared with a few others.   Well, it turns out that doesn't include the megatraps, all that's available for them is on the Megatraps page linked in previous paragraph.  It shows, per pair of megatraps, 8 Sabins reduction at 50Hz, 28 Sabins (!) between 70-125Hz.  Sabins/dollar at 50Hz is 0.016, and 0.056 at 70-125Hz.

Somewhat comparable to Megatraps are the RPG Modex Corner, it's more expensive, slight different form factor, and comes in 3 quasi-tuned versions.  Similar membrane-fiberglass design.

Real Traps brag about how Megatraps is the best.  OK, but on a cost-effectiveness they're not that much (well, a factor of 2-3x) different from acoustical foam traps, mainly because acoustical foam corner traps are much cheaper.  At most, they might be 2x or therabouts more effective per dollar spent than foam (fiberglass is about 2x as dense as foam also), and actually, the numbers I have (from promotional websites which may be misleading) suggest that some foam traps might even be better on a cost basis.

Of course, the long run issue for me may not be the cost per unit of absorption, but the quantity/quality of bass reduction per unit volume.  In the long run, there is only so much room in the room for traps, and one would want to make the best use possible of whatever space can be made available.

But I am considering the cost factor as I'm doing this very experimentally.  I don't really know what I can or need to do.  It might be worthwhile to buy some cheaper traps that I can fit now.  If I replace them later with something better, I can re-use the traps somewhere else.  So I'm looking at the cost factor to see how much I'm paying for the experiment vs how much good it might do.

Yesterday I finally started looking at foam traps.  One of the best well known is the Auralex LENRD, been around for a long time, gets good reviews, you can buy 1 for $40 and it will have between 1/10 and 1/20 the performance of a pair of Megatraps according to specs.  For the price of two Megatraps you could buy 12 or more, and as I said, it looks like that would be in the same ballpark of performance, maybe 1/2 as good, maybe not, maybe better.

Lets look at 50Hz, where I seem to have some of my worst modal problems (though that too requires more analysis), 100 Hz, and 125 Hz.

Here are the measurements on the Auralex LENRD traps. measuring a total of 48 (!) traps, they found total absorption of 59.62 Sabins at 100Hz, 76 Sabins at 125 (no measurement for 50).  That amounts to about 1.24 Sabins per unit at 100 Hz.  Per dollar that is 0.03 Sabins/$, and one could expect about 1/3 that at 50Hz (so estimate less than 0.01 Sabins/$ at 50hz).

The measurements done by Real Traps (they test a 4 ft section of 2 LENRD's) and find it has 0.52 Sabins at 50Hz.  That would be 0.26 Sabins per unit, even less than I estimated above.  So at $40 cost, that means 0.0065 Sabins/dollar, about 2.5x less effective per cost than Megatraps.  (Before I did this calculation, I had been all set to get some).  They show 5.2 Sabins/pair at 100Hz, which is 2.6 Sabins per unit, or 0.065 at 100 Hz.  So at 100Hz, Real Traps measurement is 2x what Auralex itself shows.

GIK makes Tri Traps, big foam traps for tri-corners.  According to these measurements, 8 GIK tritraps have a total of 52.2 sabins at 50hz, 98.5 sabins at 100Hz, and 102.92 sabins at 125 Hz.  Per trap, this is:


They sell for $250/2 is an excellent value, 0.05 Sabins/$ at 50Hz.  Each is 4 ft high and "2 feet wide" which probably means the open face, more like 17 inch extension on each side.

Looks to me like GIK Tritraps are one of the best values, much better value than Auralex.  A poster named Shadome at Audiogon has posted a comparison of Auralex and GIK corner traps and notes that the GIK were tested in corner, while Auralex was tested against wall, giving GIK about a 2x advantage. In his estimation, GIK is only about 1.6x more effective.  In my estimation, his estimation looks a bit low, I'm thinking corner has more than 2x advantage, so I'd guess more like 3x advantage.  (Note if he was right about 1.6x advantage, Auralex would be the better value, because GIK traps cost about 2x as much).

Anyway, this analysis suggests Auralex isn't as bad as I first thought, though I wouldn't trust it for deepest bass (50Hz) and I still think GIK is the best value...if you have room for the 17" sides vs 12" sides.

GIK doesn't actually show the side dimensions on their web page.  Here's an interesting review that includes that and other information.

Previously, I purchased These S.T.O.P melamine foam bass trapsfor about $90 each.  They don't look to be quite such a good value anymore; deep bass absorption isn't even specified, only 125Hz, for which they have 0.036 S/$.  One might expect the performance at 50Hz to be a fraction of that, 1/3 or less.  Anyway, the white color (and Class A flame resistance) are appropriate for the kitchen above the refrigerator, helping to suppress refrigerator noise (which is surprisingly high between 60-180Hz even though it sounds like a high frequency noise).

The Foam Factory has corner bass solutions, looks like performance is similar (maybe slightly less) than Auralex, but at less than 1/2 the cost.  They have several different varieties too, including the usual Auralex style wedges, blocks, and cubes, all about the same price per pound of foam.  But on further investigation, this looks like IS IDENTICAL to the very mediocre "Foam By Mail" foam that tests extremely poorly (as in not worth bothering) below 100Hz.

Tube Traps (from Acoustical Sciences Corporation and designed by Art Noxon) have been around for a long time.  They are said to be good by competitor Ethan Winer (maker of Real Traps) but just rather expensive for the performance provided, make somewhat inefficient use of corner space (otoh, very easy to install or move), and the diffusion grid concept is not as good as purpose-built diffusers.  And he says, for bass trapping, be sure to get the 20" versions.  He thinks they are fine for corners and look nice.  He thinks his own panel absorbers are a better choice (and far cheaper!) for flat walls.  ASC on the other hand sometimes doesn't put them in corners, but creates whole walls of them.  Note that such walls of Tube Traps can be astronomically expensive, one guy spent $38,000.

Now that I see that they are completely sealed (cardboard surrounds fiberglass, then further inclosed in structure and cloth) I feel comfortable with putting them near HVAC intake.  They may actually be better "sealed" than Real Traps (!) which rely on cloth cover to prevent fiberglass leakage (though there is also a "membrane" which prevents that also).  I suspect in either case, actual fiberglass leakage may be less than what is simply tracked into the house from the garage.  Ar

Here is Art Noxon's original AES paper describing the design of Tube Traps.

The price is formidable indeed.  A 4 foot 20" round tube trap costs $854 on the factory direct pricelist.  As a 4 foot high corner section, that's equivalent to the 2 GIK Tri Traps you can buy for $259, or the pair of Megatraps you can buy for $500 (though most likely the Megatraps have the highest performance of these options because they are the biggest...almost value comparable with the GIK...the GIK still looks like the best value, especially since I can't see how I could fit the larger units).

What would really be cool, IMO, would be an 8' tube would fill up the whole corner (like by my HVAC intake) pretty nicely (though...even with curvature there would be some blockage to HVAC intake.  The price for taller models isn't listed, one can guess it's probably at least twice the price of the 8 foot section, so $1708.

The best option for that corner might be to forgo trapping on the very bottom, then start trapping just above the HVAC intake.  Thus there would be no blockage right in front of the HVAC, but there would be blockage of intake *above* the vent, not sure how crucial that is.

Another option would be to have tube trap in the bottom section, then GIK on top, less blockage below.

One idea occurs to me now: cover up existing HVAC intake and turn water heater (beneath HVAC) door in kitchen into new HVAC intake.  (I've now spent days thinking about this.  It *could* work, there is plenty of space on door for needed cutout, unless done very skillfully with some kind of wood or metal grille, it would be very ugly.  I will probably need cabinet maker to do the needed work.

Another company that makes traps is Primacoustic.  Their bass traps include two thicknesses of flat panel, a solid foam wedge (nicer looking than Auralex but also way more expensive), and a flat triangle tri-trap that looks lighter than the Real Traps version, but probably easier to mount.

As a unique solid (not sculpted) foam, the Australis looks especially interesting to me.  I think solid foam should be better than sculpted for low frequencies.  (Though, one could also get Auralex cubes...)  You get 6 linear feet of corner damping for $299.  You get the much bigger fiberglass-based GIK traps for slightly less, and get 8 linear feet.  So the Australis does not look like the value leader.  But the small and relatively light traps can easily be hung...they are expressly designed for hanging while the GIK are designed for stacking.  Though the specs suggest the Australis works competitively, a 12" extension foam trap just cannot be as good as a 17" extension membrane/fiberglass/airgap/panel trap.

The beauty of the Australis is it's easy mounting.  You can always use them somewhere, even at ceiling wall corners.

Wednesday, February 23, 2011

Weak Bass

Listening to DSOM, Supertramp COTC, and Dire Straits Brothers in Arms, they don't sound bad, but just a bit weaker in the bass than I'd like, and not as "sweet" as I like.

I then listened to Brothers in Arms on Bedroom System.  That system really rocks when you are sitting at the headboard, which is at a wall boundary.  It's very sweet and surprisingly transparent ($2199 Revel M20 monitors which IME far surpass LS3/5A and other classics, though comparable to modern delights such as Harbeth M30 which are even more expensive), though lacks the ultimate spaciousness of the dipolar living room system, not as much difference as between being in a movie theater and watching a 32 inch TV, but going in that direction.

It appeared I could make the bass in living room system more full if I move the listening spot backwards, even just 10 inches to where I was sitting last year.  But the problem with that is it reduces the stereo separation, which I have come to value more strongly this year.

I suppose best solution would be to move both speakers AND listening position further away from the front wall.  The cancellations at mid-room are most likely room modes and are associated with the listening position in the room, not the speaker position in the room.

That's just not very practical.  My multipurpose living room is already being taken over by stereo equipment, most people would consider it bordering on some kind of insanity.  The speakers are already the recommended 3' from the back wall, taking up a rather significant chunk of floorspace.  Further movement of the speakers runs smack into the entry path (would have to move speaker to get into the room) and the couch.

I think I will go back to trying Tact room correction again, but following the strategy of setting the "room curve" to pretty much match actual measured response, except with a slight tilt away from the highs and toward the midbass (deep bass already loud enough thanks to room and corner gain).

Maybe that would be reinventing the Quad preamp "tilt" control.

I wasn't happy with my collection of 3 and 4dB boosts in the midbass that I worked out last weekend.  They seemed to cause some sort of compression like near clipping.  And they didn't seem to help much either, bass sounded equally weak either way.

I suppose there is also the "move speakers out when listening" strategy.  That sort of thing always gets in the way for me, I wouldn't listen as much, etc.  I like systems that are ready to play at a moment's notice, and moving speakers is alway a job that takes me about 15 minutes because I'm fairly obsessive about positioning to nearest mm.

This isn't really dependent on subwoofer level, the midbass weakness is mainly in the panels which now play from 84 Hz up.  Measurements suggest weakness from 60Hz - 250Hz. (Detailed measurements show a series of 4 10dB cancellations, which I tried to doctor unsuccessfully with EQ.)

I do plan to do such things as measuring the sound at potential listening positions.

The in-wall metal fireplace is clearly having a bad effect also.  Actually I now believe I can convert the fireplace into a kind of damper by lining the walls with heavy magnetic vinyl (i've got several rolls) and then stuffing the inside with the same fiberglass as is used by bass trap makers, or something like that.  The real coup would be making a "membrane" damper by putting some lossy thick plastic diaphram over the front to convert pressure to motion. Membrane dampers can be quasi-tuned also.

(As an experiment, I'm sending this both as an email and a post to my blog.)

Monday, February 21, 2011

Rattle found

It wasn't anything on the mantle that was rattling...  It was mother's picture.  I recall that rattled when it was in the back corner of the room also.  It seemed to go away with just slight movement.  But I think it needs either better damping (foam pads on back) or relocation.

Glad it wasn't mother's urn.  That can stay on the mantle where it belongs.

Very thorough analysis of audio system issues

For example, they analyze the effects of different interconnect impedances and determine the optimal source impedance is in the neighborhood of 50-100 ohms.  Actually 90 ohms yields the lowest peaking (most interconnect systems peak at some ultrasonic frequency; it's LCR with very low R) whereas 60 ohms yields the most extended response (from the low pass network created).  Current best practice in audio is balanced interconnection, 50-100 ohms source impedance into 100k ohm load, 200khz bandwidth per component (in a multicomponent system, all the roll offs add up, so you need margin to maintain 20khz, they assume 5 analog components in series is possible), and 5v/us minimum slew rate (less than that, and you get slew rate limiting into the now pro standard +28dBu (aka 20 volts RMS) at 30kHz.  (Then they laugh at adage "a system is only as good as it's weakest link", of course a real system is always *worse* than it's weakest link.)

Clever idea to fix bass null electronically

How to fix the 46Hz null in the left channel electronically (!!!)

My clever idea is this...playing with the fact I have dual subs.  I'm using the subs in "IRS" model (like the old Infinity IRS speaker) where each subwoofer gets the low passed signal of one channel.

Actually, lots of Home Theater hacks say you should run the subs in mono if you're crossing over below 100Hz.  I could try that.

Or I was thinking of some kind of complex crossover, where the bass is steered between one sub and the other in order to defeat the null.  Actually, there is no null on the left channel, so I could have a second crossover that adds puts all the bass below 50Hz on one side.

Right sub: 50-85Hz
Left sub:   15-50Hz (mono) + 50-80Hz

Or, for even more fun, just cross-over the 35-50Hz range:

Right sub: 50-80hz + 35-15 hz
Left sub:    50-80hz + 50-35hz (mono) + 35-15hz

By now, I hope you are laughing (and not just shaking your head).

I'm don't think my digital crossovers can actually do that, I might have to use two of them (I have a spare).

What measurement microphone to get

One of the reasons I downloaded the freeware program Room EQ Wizard is that one of my biggest issues wrt buying RplusD is whether or not I should order their calibrated microphone.  They say that it's not necessary for most applications.  I already have a bunch of measurement microphones and SPL meters of varying quality, and among them several calibrated ones.

But, what I really want to do is get a Really Good Microphone, like an Earthworks 40.

Anyway, with the freeware program I can test (somewhat) the overall configuration without feeling pressure.  (I coulda also downloaded RplusD for free trial... I do that next.)

I did get everything set up using the Galaxy meter.  For a better test, I want to an actual phantom powered microphone, like one of my Behringer ECM8000's (which are hard to find right now, buried in the junk, I have been using only Tact microphones for past couple years), or one of the Tact's.  The Tact microphones are calibrated, or at least they come with calibration curve done by Tact (which some people think isn't very good), and I think they are slightly higher quality than the ECM8000.  But I'm not sure if they can stand the 48V phantom power that the Emu Tracker Pre puts out.  The Tact only seems to supply 10V phantom power.

Anyway, the Earthworks would do fine on the 48V.  If the Behinger works (however good it's response is) the Earthworks should work too.  The Behringer is rated for 15-48V phantom power.

Anyway, REW is said by Ethan Winer to be more useful than ETF in some applications, though his main recommendation is ETF (the predecessor to RplusD, so I think he would recommend the later product today).

Fix that rattle!

After recent adjustments, on Sunday I though I'd go back to play "Spanish Harlem" sung by Rebecca Pidgeon on the "Raven" album.  That has a bass line which has been said (by some famous audio engineer) to be good for tuning bass.  The acoustic bass notes should be roughly equal in volume.  (Actually, on most of my systems, despite having subterranian bass, I get the third note, the higher of the first three, to sound slightly louder and different in character, as if it's a plucked open string, though I'm not sure that's true.)

Anyway, now I noticed some very annoying rattling on that third note.  Thinking it was a panel problem, I muted the acoustats.  Now the bass line coming from the sub is transparent through the entire piece (that's all I hear on the sub) and the third note still rattles horribly.

I tracked it down to the fireplace mantle.  ASAP, I'm going to start removing stuff from the mantle until I can make the problem go away.  Now that this problem has made itself very clear, it's a good time to fix it.

Sunday, February 20, 2011

Room EQ Wizard running

Tonight I've downloaded and installed the highly praised freeware program Room EQ Wizard (REW).

Also set up Emu Tracker Pre soundcard, and Galaxy CM-140 (the SPL recommended by the makers of REW).

Just putting SPL on top of listening chair for now, and didn't get CM-140 calibration downloaded yet.

Measurements are not showing a peak at 42 Hz.  They are showing, at least for the right channel, a deep null at 46Hz.  (The peak from 38-55 Hz is in the corner, not at listening position which gets the opposite response.)

Left channel showing 10dB dips at 125 and 165 Hz, so I reset my EQ's for that, with 4dB rather than 2.2dB gain.  Even though I set octave width at 0.25 for 125, the EQ simply seems to raise the entire region.  So I changed the octave width to 0.35 for less ringing (presumably) makes very little difference.

UPDATE: A sweep up to 2kHz reveals that the largest/widest suckout was really around 230 Hz, so I added an EQ for that.

Current EQ's
42Hz       -2.0dB 0.4 octave
125Hz      3.0dB  0.3
165           3.0dB  0.3
230           4.0dB  0.35

Result looks like an improvement, anyway.

Friday, February 18, 2011

RPG Modex Corner Bass Traps

I've found a different bass trap, this may be the best for my primary resonances 38-50Hz.  This is the RPG Modex Corner.

This seems similar in concept to Real Traps Megatraps, but RPG makes them in three different tuned versions, 40Hz, 63Hz, and 80Hz.  What I need mostly is *SERIOUS* bass trapping around 42 Hz.  The remaining problems are trivial.  Funny, the RPG website doesn't mention that the unit comes in 3 tuned versions.  These are not, of course, pure Helmholtz resonators.  But if the "tuning" optimizes absorption of deep bass (which is really hard) that would be very useful.  Most traps don't do much at 40 Hz.

Following DIY discussion on these things, the consensus is there is no good way to predict response, even very knowledgeable designers have to cut and try.  Heavier membrane yields lower cutoff frequency.

Acoustic Treatments

Here's a great article on room treatments by Ethan Winer (maker of Real Traps). Ethan seems like a very good guy and I trust him.

Looking at my room though, it's hard to see where I would fit the Real Traps.  I have a very nice spot for a 16 or 20 inch diameter tube trap, but I'm not sure I like the way Tube Traps are constructed with fiberglass, which I'm worried about leaking through the fabric and into my HVAC system.  Real Traps tend to enclose the fiberglass in plastic, which seems safer.

I'm planning to order RplusD, the premier acoustic measurement software, today.

I'm still liking my manually adjusted EQ's, though I really want to measure and "improve" them.  I listened to some Bach organ works today and was delighted by the clarity of the bass.

For measuring my manual adjustments, however, I have to dial them into the Behringer.  (Currently I dial them into the Tact Parametric EQ, but that is turned off during measurement.)  So it's a bit of a pain, and the Behringer can't be adjusted with as high resolution as the Tact.

Thursday, February 17, 2011

Acoustic measurement programs

Ethan Winer (Real Traps) suggests one of the following programs:

ETF  (now replaced by RplusD)
Room EQ Wizard
Fuzzmeasure (for Macs)

Here's a blog where he mentions these and links to his instructional information.

Bass Traps

Recent experiences have suggested to me that Room Correction EQis not only far from panacea, it's not clear if it's worthwhile.  Indeed, last year when I had corrected/boosted response at the listening position I had so much boom elsewhere it was causing lots of unnecessary rattling and made guests uncomfortable.

Meanwhile, I do like to do small EQ adjustments by measurement and ear, those actually do seem to help and don't do much harm.

So count me as a new believer in the need for room treatment, especially bass traps.

Leading bass trap manufacturers include Tube Traps (cylindrical and quite expensive for the larger ones), Real Traps (somewhat less expensive, look like better value), and others.

It will be a big problem figuring out where to put these things, the room is already quite stuffed.

OK, just the basic EQ's now

OK, until I can figure out what I'm doing better, I'm a bit worried about using high Q peaks to fix dips as I was apparently doing with some success on Tuesday.

So I'm back to making this optional EQ #1 in Tact:

Cut at primary resonances:   -3dB 42 Hz  (this is very conservative adjustment, could use lots more)
Small boosts to midbass" +2.2dB at 125 and 140 hz
NEW: small cut to high end, -0.7dB at 13.7kHz

Sounds quite nice this way.  The new small cut to the high end seems to remove high end glare on Shadowfax "Nuclear Village" album.  No EQ sounds fine until you try the EQ, then you don't want to go back.  If high end cut is increased to 3dB, however, it sounds too dull on top.

Meanwhile, I was thinking why automated Tact didn't work.  Tact seems to show Acoustats peaking in room at 18kHz.  Other measurements with SPL meters and sinewaves suggests the Acoustats peak at 13.8kHz and fall off above that normally.  I retested that using ST 1100A distortion analyzer as my sine generator and Galaxy 140 SPL meter and GR 1933 SPL meter.

GR is very hard to interpret because being random incidence.  The Galaxy shows about -3dB at 15kHz and -6dB at 16kHz, definitely does not look like 18kHz peak.

However, unfortunately, Galaxy is only Class II rated to 8kHz.  The rolloff I am seeing could very well be rolloff in the meter itself.

Or it could be, as I've suspected, that the Tact microphone isn't calibrated correctly.  If the calibration, for example, shows -12dB at 18kHz, but the mike is actually flat, it would add a high frequency peak at 18kHz in the measured response.  On Tact user's group, complain about mike calibration is legendary, a standard recommendation is to scrap the manufacturer's calibration and simply substitute a generic calibration for the mike.

It's also very hard to measure with high frequency sine waves.  Tiny changes in position or angle can make 6dB or greater difference.  At each measurement, you need to move the mike around a bit to find a solid reading.  It would be nice to have warble generator.  However, the best of all would probably be RplusD from Acoustisoft (formerly known for ETF).  Can get whole package including calibrated mike.  I think I will do that this month.

Tuesday, February 15, 2011

High Q Parametric Corrections?

I've found that while I can correct a small depression around 60Hz, I need a very high Q filter to do so nicely.  Q=8.9 and gain=+15.  Without the high Q, the boost simply moves a whole region up, with the nearby peak moving up nearly as much as the depression at Fc.

I'm worried that such a high Q filter doesn't really correct the response so much as swamp a local region with a Very Big Peak.  This is not visible on the Tact measurements, but I think a finer grained measurement, like one from AcoustiSoft, might do so.

Alas Room Modes are not "minimum-phase"

According to this argument, many electronic circuits are minimum phase, and loudspeakers have both minimum-phase and non-minimum-phase effects (with minimum phase characteristics dominant, it is claimed, in the better loudspeakers).

Rooms, alas, have delayed resonances, the epitome of non-minimum-phase.

So what about this... I could in principle correct the speaker minimum-phase effects *perfectly*.

Room effects, I can only broad brush, which may help or harm.

Some of the fancier room EQ systems now measure both speaker and room effects (using gating) and correct the speakers first.

It's not exactly clear what my Tact system does.  The description of what it does does not reveal all the technical details, and the system is not open except to the extent that you can edit a target curve.  Reviews are mixed but most say it's better to have target curve capability than none, so it's better than most such systems.  Most consumer room correction systems have preconfigured target, they are pure black boxes, take it or leave it.  Now there are systems which brag about their approaches in more detail, and may have more configuration options than just setting a target curve, but they tend to be really expensive.

There are open source solutions to room correction by EQ, so you can examine or change the algorithm, but the most developed one works with an obsolete Behringer feedback canceller (has lots of parametric EQ's) because it's cheap.  It has hum problems, uses semi-pro rather than audiophile line levels, uses a low-ish sample rate.  I bought one at close-out, but then was lured in by the much nicer quality Tact hardware.  The open source solution auto-programs the old Behringer through it's serial port.

You could say that any-old computer could do the trick.  But any-old computer has a noisy power supply, fan, etc., you don't even want one in your listening room.

In principle, given a system with 10 or more parametric EQ's, you can still use the open source software to figure the optimal parameter settings, then just toggle them manually into Tact or DCX crossover.

I have 13 manual parametric EQ's per channel available in the Tact and about the same number additional ones available in the DCX.  Not quite the same as the hundreds of poles Tact can use through it's automated room corrector, but enough to do some good or considerable harm.

Monday, February 14, 2011

Manual parametric EQ instead of Room Correction

I have been disappointed by this round of Room Correction with Tact.  It very much looks like I cannot do a decent room correction without setting up a target curve which is pretty close to my measured response.  I suppose I could do that, but I am not so motivated anymore.  It is clear that the correction I used last year was leading to excess room boom at 45 Hz and other problems.  Now I have heard how a simple correction can lead to very undynamic sound.

Instead, I am doing what I have long done with my bedroom system, applying manual parametric EQ adjustments.

Starting with no EQ or RCS, I do find that the primary room nodes 38-50Hz are turned up slightly (though this is HIGHLY position dependent, and one can even get cancellation nearby).  So a manual EQ should cut this back slightly.  I am using the 6dB cut of Tact Parametric EQ #1, but maybe should roll this back a bit since it may cut too much at the listening position.

Also, a manual EQ should attempt to improve the 100Hz-200Hz depression.  I think at least part of this depression is caused by frontwall cancellation, also maybe partly by dipole cancellation.  My strategy for fixing this is to use multiple staggered small EQ's.  I have started with 2  corrections:

140 Hz  +2.2dB  "octave width=0.6"
125 Hz  +2.2dB  "octave width=0.3"

(I have programmed these into Tact EQ #1.  The Tact allows you to select "octave width" instead of Q.  It is nice to be able to select EQ's by remote control.)

The above EQ's have about the right effect, increasing the 100Hz-200Hz response for a pleasant, if not perfect, correction, in combination with an EQ at 42 Hz.  The attempt is to make a broad pleateau in the electrical response from 110-180Hz, with the plateau getting taller as we get down to 120Hz (but avoiding peak stimulation of the potential resonance area at 116 Hz).  It is often said you cannot correct deep depressions with EQ.  That is true, but you can ameliorate small depressions.

Using these EQ's, the outside of the listening area has much less "boom" than last year, but there is still some boom in the corner near the Kurzweil.

With the EQ's, the sound is very sweet.  It is one of these paradoxical effects that improving the midbass seems to improve the mid highs.

Testing the Acoustats

It's interesting to listen to the Acoustats playing through their crossover without powering the subs.  This way you can hear more clearly exactly what the Acoustats are doing.

With 121Hz crossover, the Acoustats sound like a tweeter.  Zero bass.

With 84 Hz crossover (now LR48), the acoustats sound a little thin but still like a full range speaker.

I brought up two versions of Polyvtsian Dances on Friday Night, and played them as loudly as I could with the new notch filters in place.

No buzzing or rattling whatsoever.  It's looks to me like the 84Hz crossover is safe.

On Sunday I tried to duplicate the buzzing problem with "A Story Within A Story."  I couldn't make it buzz anymore, even with the notch filters turned off.

Perhaps by operating the Acoustats with a lower crossover, they have "broken in" and loosened up so they don't buzz anymore at 116 Hz.

Actually I should probably go back and test Polyvtsian Dances without the notch filter too.

It does seem that the octave from 100Hz to 200Hz has pretty weak output, particularly from 110-170 Hz.

Taking out the notch filters doesn't change much (surprisingly) but does make the output just noticeably weaker.

I had been motivated to replace the notch filters with "dynamic EQ" filters that activate only for signals loud enough to cause the buzzing.  But I can't know how to set up the dynamic filter if I can't reproduce the problem it is supposed to fix.

New Supertweeter Adjustments

I've been worried about the high supertweeter level (+15dB) actually causing some problem in it's own right.  Normally SPL levels above 20kHz are very low.  But you could have some ultrasonic noises, and even at -15dB such noises could cause huge signals in the supertweeter drive.

Something like that seemed to be happening when I played the song Atom Heart Mother Suite by Pink Floyd.  The supertweeter channels appeared to be clipping as the red lights were lighting on the Behringer.

So I spent the next 12 hours working on adjustments to fix this, only to later find myself unable to reproduce the original problem to see if the adjustments really help.  The original problem might have been caused by some kind of gremlin (like loose connection, or maybe I mistakenly turned off LR48 crossover for supertweets).  But I still think the adjustment is worthwhile and makes for better sound, so I keep it anyway.  (I had been thinking I would try dynamic EQ instead, but since I can't reproduce the original problem anymore, I'm sticking with the new changes.)

Mainly what I did was add a low pass parametric EQ at 20kHz in the supertweeter drive on the Behringer.  I choose 12dB/octave and -15dB level (I'm not sure why Behringer requires you to set levels for a low pass; I think it actually makes it like a shelf rather than a low pass.)

So I have both lowpass and highpass at 20kHz.  And I think this *is* the right way to set up a supertweeter: with both highpass and lowpass.  (The highpass is a Linkwitz-Riley at 48dB/octave at 20kHz).

With just the LR48 highpass, the level at 20kHz is actually 6dB lower than the maximum (which it approaches asymptotically above 20khz, being pretty much there at 25kHz, just as could be predicted).  Now what good is that?  Perhaps the supertweeter needs the extra 6dB boost above 20k to maintain flatness, though I would think not.  But whether it needs it or not, that is just the way using LR48 works out.

And it's highly undesirable to have the 20-40kHz range be boosted that high, it's just asking for trouble from ultrasonics, etc.  What we'd really like is response that keeps on rising until 20kHz, and then doesn't rise anymore, maybe even falling back slightly.

Just to get a sanity check on the voltage sent to the supertweeter, I put my Fluke 8060 and 804 meters on it, and drove my system using my Sound Technology ST-1400B distortion analyzer/generator.  I set the level for about 1V output at Acoustat speaker terminals.  At 20kHz, the Elac drive wire was about 2.5 volts, and it peaked around 30kHz at 5 volts (!), 14dB higher than Acoustat drive level.

After adding new 20k lowpass, the peak occurs near 20kHz at about 1.8V, then declines gradually.  I also changed crossover point for the LR48 high pass to 18kHz.  Although I had recently changed levels down to +10 on the Behringer, I am back to +15 with these adjustements (which lowered response quite a bit by themselves).

Not only is this "safer", I think the supertweeter sounds better not being driven so loudly as it was.  That's exactly when it makes the "metallic" sound.  The new highpass/lowpass combo seems to prevent that.

Symphony SPL levels

Last year I measured the San Antonio Symphony at 101dB.  A friend questions this.  Here is a link to other people talking about SPL levels, one guy measured Minneapolis symphony at 106dB.

Friday, February 11, 2011

Sometimes it's best to do things the right way (LR48)

In earlier post I described various strategies for crossing over (or not) my electrostatic panel speakers and my subwoofers.  Last night, I found that one of the strategies I have passed over for awhile does indeed work very well.  That is the strategy of having a symmetrical crossover, with high pass and low pass cutoffs set for the same frequency and using the same crossover function (such as Linkwitz-Riley or Butterworth) on both sides with the same ultimate slope (such as 24dB/octave).

Listening to some FM radio for the first time in a couple months, I was first impressed by the clarity of the sound and incredible spacious imaging.  (This mainly results from moving the listening position forward about 8 inches, also from recent changes to the sub/panel crossover.)  But I was also appalled by an apparent dryness, bordering on harshness.

Now I've been fiddling a lot with the crossover controls recently, but when I made the above observation I had the panel highpass at 104Hz with BU24 (butterworth).

First I tried changing that to 84 Hz, same as the lowpass on the subs.  Big improvement, now the sound is sweet and less dry (funny how improving midbass makes the midrange sound nicer, but not surprising really).

But then it actually sounded a bit boomy.  So I changed the BU24 to LR48 (Linkwitz-Riley) and that cleaned up the bass.  Now it was sounding very nice and I just kept on listening for quite awhile.

Also funny that when I started and was thinking the sound was harsh, I turned off the humidifier like yesterday.  But that didn't seem to have as much affect as changing the crossover.  After changing the crossover, I was enjoying FM radio with the even noisier dishwasher running.

Now the idea that both highpass and lowpass sides of the crossover should have cutoffs at 84Hz, and that they should both use LR48, that sounds pretty obvious, right?  So why wasn't I doing this already?

Well, with somewhat less careful recordkeeping, I did try this last year.  But one of my goals way back then was to take advantage of my dual subs and improve my ability to play more loudly by offloading bass from the panels (which can't play back much bass without bottoming) and onto the sub (very high output capability).

And I had a test, one version of Polyvtsian Dances, where a tympani made by Acoustats rattle.  I found that I could get rid of the rattle by moving highpass point to 121 Hz.  So that was where I set the crossover and where it stayed for about 12 months.  (Actually, for other reasons, I set the sub lowpass even higher, and then let the Room Correction fix up the response, a very bad approach I now know.)

Well I haven't repeated the rattle test, but now I have a notch filter at 116Hz specifically to handle the rattle.  So now I may be able to have good sounding bass without the rattling.

The other point was that there are terrible room modes in the range of 90Hz-110Hz.  I could in effect EQ these down by spreading the sides of the crossover apart (I call this underlapping).

It turns out to be a very bad idea to have sub playing in the 90Hz and up because it excites room modes horribly.

But it works out fine to play the electrostats down to 84 Hz, because their dipolar character basically doesn't excite those full-wave-and-up resonances.

Hearing how much better it works to have full crossover at 84 Hz, I don't think I'll be trading that away for loudness capability any more.  Except I may have a secondary setting for those days when I need more loudness.

I have also been thinking about the possibility of making my notch filters amplitude dependent.  The Behringer does provide that option, but it might do more harm to the sound than good.  But along those lines...why not have amplitude dependent crossover point?  When you play really loud, the crossover automatically ratchets up to a higher frequency to protect midrange panels.

Fully crossing over both sides with LR48 how does the bass sound?  Almost like pure electrostatic.  Interestingly, LR48 like all LR crossovers actually reduces energy around the crossover point (helping to avoid feeding those resonances) but maintain amplitude in the axial response, so that straight ahead the drivers add up arithmetically.

Thursday, February 10, 2011

Strategies, large and small

To move forward in the audio hobby, you should have a strategy.  There are many possibilities nowadays. Unfortunately, it looks like a lot of the current audiophile products are of the "fake snake oil" variety, they don't actually work, but people feel they work because they believe in them.  The tendency toward superstition in humans is very well known...

My strategy is largely to avoid things that look or smell like "fake snake oil", or at least not to get overly obsessive about them.  Sure, I may try a tweak here or there, something that makes sense to me like power conditioning or replacing electrolytic capacitors with polyethylene film capacitors.  (I am well aware that power conditioners are quite controversial, and in fact I do not plug my power amplifier into mine, only signal processing equipment.)

Instead, my strategy is to focus on the well known, things that unquestionably make a difference, no question about audibility, but which is better?  There are vast opportunities of this nature in audio, and I would like to bring more into thinking about audio reproduction like I do.  I can see that I am not alone, however, lots of board now (such as DIYAudio) are filled with audiophiles having a more objective orientation.  That does not mean I treasure flat frequency response above all else (see my last post).  But at least I want to have some understanding of the colorations I employ to make the audio magic work.

It is clear to me that a very important part of this is getting the bass right.  Mind you, I don't believe it's entirely clear what exactly "right" is, and it may vary from one person to the next.   It is clear to most people who have investigated this that the best sounding frequency response in a room is not flat.  Rooms themselves have low frequency resonances and gain, and within that context if you hear "flat" reproduction it really falls flat.

So this is a subjective art, basically fiddling with the bass until you get it to sound right on most recordings.  Now there are many approaches to that.  Some people go through many different speaker systems, and/or try many different speaker and listening positions with those speakers.  Others stick with favored speaker and try room acoustical treatments or DSP.  Still others fiddle with magic points and tourmaline crystals.

I actually have been through many speaker systems in the past, and I think that is one of the best things to do, but it is not my current approach because I already have very good speakers.  Currently I am mainly trying to get the bass right by adjusting the speaker positions, delays, crossover, and digital signal processing.  It is clear now I may also benefit from some room acoustic treatments to fix room boom not at the listening position.  It's clear that neither DSP doesn't offer a complete solution to getting the best bass response.

So now I am getting down to the layer of strategies where I am now.  Since January, I've been working with a new crossover design that keeps the subwoofer below major resonances around 100 Hz, and the panels above those resonances.  This works very well, though seems a bit lightweight in the midbass (yes, it seems light around 100Hz even though it measures basically flat).

Now there are several other possible crossover strategies.  One is to do a more honest ("real") crossover, with subwoofer lowpass and midrange highpass both set at the same frequency, say, 100Hz.  That will likely produce lots of "boom" around 100Hz which is currently being suppressed.  It wouldn't be a nice way to use the system without further changes.  Those additional changes could be either done manually, by setting a parametric EQ depression around 100Hz to cancel the boom, or by running Tact do do a Room Correction, or both individually or simultaneously.  But those approaches may work out better than trying to cover up the resonances by underlapping the crossover.

Another obvious strategy is to run the Acoustat full range without sub.  Well not if you like head banging bass as I sometimes do.

Another strategy is to run the Acoustats full range with subwoofer merely serving in augmentation mode.

Similar to that would be to crossover the Acoustats as low as possible, such as 40 Hz.  I tried that before, didn't like some of the panel resonances I heard running that way, but those may be the very resonances I fixed with notch filters this week.

Yet another is to run the sub as high as possible (it's supposed to have flat clean dynamic response to 250Hz) and cross the Acoustats at that point.  That would keep as much bass out of the Acoustats as possible, and give the Tact as much freedom as possible to correct the frequency response (which will definitely be needing lots of correction in the midbass when powered by dynamic ported woofer).

I think for now I'm going to stick with the crossover-around-100Hz strategy, and work it up to get better midbass, or attempt to get some satisfaction with the "real crossover plus EQ" approach.

Better without RCS (Room Correction) ???

Another late night with good listening and important audio discoveries.

1. Midbass better with Butterworth 24 than Linkwitz-Riley 48.  Even with the room curve(s), which currently boost deepest bass below 35 Hz and highest highs above 18kHz, there seems to be a broad depression in the bass using the LR48 crossover for the Acoustat high pass.  BU24 gives much nicer midbass sound.  By mistake (probably) I had originally done this series of room curves (from Sunday February 6th) using the BU24.  Based on the impulse response mainly, I had decided that LR48 was acceptible (and preferable for getting the buzz causing bass out of the speaker).  But then I went ahead and ran the RCS measurements with the BU24 selected anyway, because that was what I had just finished testing.

I had determined by Monday that I had made this mistake, but decided to switch back to LR48 for speaker protection and (thinking Acoustat response below 104Hz crossover is nothing much to speak of anyway) thinking it wouldn't make much difference.  But all along I have been noticing weak midbass around 100 Hz and just below (which is party by design...underlapping to crossover to minimize the effect of bad resonances around 100Hz).

But until I installed the notch filters on Tuesday, I really did want to keep the buzzes out, and the weak midbass was a small price to pay.

But now I am using notch filters to get around the buzzing (and it continues to work great), I shouldn't have to compromise the rest of the midbass as much.  So I went back and tried the BU24.  BU24 brings back much of the missing midbass.  And actually it does this regardless of whether room correction is used.

This many not be an inherent property of BU24, but a set of circumstances in my system and listening room.

#2 Currently, the uncorrected sound is better ???  Well now that I have decent sounding system without correction, I can run system either way.  There is a huge difference with my current correction curve #2 (which actually had its room curve tweaked on Tuesday) and bypass.  #2 reduces the highs (which are generally too hot, I admit) and brings up the midbass (only slightly).  It clearly reduces resonances, you can hear much more detail in the music because of the reduction of spurious resonances.  But by comparison with bypass, the corrected sound is flat, undynamic, and boring.  (You don't notice this until you actually compare the two.)

I don't think this is entirely a situation of preferring the euphonic colorations of my system.  I think it mainly stems from the fact the the frequency response of the Acoustat, with its tilted up high end in axial response, is a deliberate design choice to make up for the relative beaming of higher frequencies.  And I may have to do the correction of bass and midbass better with a better chosen room curve.  I will try something closer to actual measured response.

I may be able to make the correction sound better than bypass with addition changes to the room curve or multiple measurements.  But for now, I like bypass better.

#3 I can now play Supernatural by Santana (SACD) without unpleasantly strong bass shaking the room apart.  Big improvement!  It sounds like an entirely different and much better recording now.  It was a big disappointment last year how badly this recording sounded without an extra bass cut.  Now I can listen to the record without even engaging my "boom correction" filter.

#4 Best to shut humidifier off when playing.  My noisy humidifier is about 10 feet from the listening position.  I run it all winter to keep my throat from drying out.  But when I was playing Dark Side of the Moon tonight (in great contrast with last night) I found the side slightly on the harsh side.  The sound generally seemed to improve (though I didn't go back to DSOM) when I turned off the humidifier.  Duh!  But the big problem in situations like this is how to remember to turn humidifier back on.  (It was even worse when I was shutting off refrigerator.)  But now I have a trick.  I turned off the light in the kitchen where the humidifier is, and shone a tensor light on the humidifier.  This trick almost didn't work, as I first went to bed without looking at the kitchen, but I caught it on my first trip to the kitchen a few minutes later.

#5 Supertweeter back to +15.  It certainly doesn't seem to add to harshness, not sure if it can actually reduce apparent harshness, but it's usually more fun to have supertweeter at higher level, and with 20kHz level, I think it has lower output than last month (when I was using 15.5kHz LR48 higpass, now I am using 20kHz BU24 highpass, higher but less steep crossover).  Not sure yet if I want to keep it like this.  I also seem to like room corrections where I leave supertweeter off for measurement, then add it back in for measurement, as I do with correction #3, but #3 currently doesn't enjoy the additional room curve tweaks I put into #2.

#6 One usually thinks that the panels are producing great bass.  But just turn the subwoofer off, and it's clear that the panels are mid-tweet drivers.  With my crossover anyway.

Wednesday, February 9, 2011

Why is there uncorrectable room boom in the first place?

I have a fairly typical sized living room for USA, it's about 14 by 15.5 feet with an 8 to 9.5 foot vaulted ceiling.  The passageways through entry and kitchen are wide so they also may make the room slightly larger acoustically.  My primitive calculations suggest I should get half wave resonances in the range of 30-40Hz, however they actually appear around 38-55hz.  It so happens that my Kurzweil synthesizer is right in the back corner of the room where you get a hugely resonant response (body shaking) in the 38-55hz range.

Now one interesting thing about half-wave resonances is this: In a closed room, they have their maximum amplitude away from the center of the room and towards the walls.  Full wave resonances have a third maximum amplitude in the center.  My full-wave resonances should be at double the frequencies of the half-wave resonances, so about 76-110Hz.  (I have dealt with that part of the boom by underlapping the crossover and by applying room correction, as described in earlier posts.)

So it seems, probably, that typical rooms have half-wave room boom in the range of 30-60Hz.  That's a pretty important band for bass fundamentals.  Half-wave boom is the kind that seems to accumulate around the walls of the room.

Now most serious audiophiles listen to music from a listening position more in the center of the room rather than the periphery.  If you do that, and you have half-wave boom in the range of 30-60Hz, you could probably benefit from the Boom Control I described in the previous post, to handle recordings with a lot of 30-60Hz energy.

Here's an interesting page on standing waves and resonances...I need to learn more here...perhaps my analysis isn't entirely correct yet.

Thinking about the above, it appears that the room modes (1/2 wavelength, 2/2, 3/2, etc) indicate the room is operating like a open tube because the maximum volume (antinodes) for those are on the sides.  The quarter wavelength series works like closed tubes, and typically have node on one side and antinode on the other.

Here's the Wikipedia entry on Room Nodes, which notably includes the comment that the attempt to equalize the sound in one position may actually make it worse in others.

Here's another interesting discussion showing another room effect, the 1/4 wavelength from the rear wall effect.  Also nice ETF measurements of actual room nodes:

my latest invention: Boom Control

By now, it's pretty clear that using digital signal processing (DSP) both to analyze my system and correct it are very useful.  I'm really enjoying the new cleanness that comes from having used notch filters to negate some very serious buzzes in my Acoustat panels.  I imagine now I was hearing those buzzes all the time, and I thought they were coming from something else, like the subwoofers.  I'm worried that the 116Hz resonance may not actually be easily fixed by giving the panel membranes a hair dryer heat treatment.  It may be something related to the frames or individual panels, and a repair may require added bracing, etc.  Lots of people find they have to brace their Acoustat 1+1's for better sound.

But moving on, DSP also has limits.  Flattening the frequency response at the listening position may not remove serious resonances (room boom) elsewhere in the room.   Now, you could take the position that these resonances don't matter.  But they often add a "strained" quality to the sound, as you hear walls shaking, etc.

This is an old argument.  Those who sell acoustic treatments often say DSP is worthless, and vice versa.  The truth is that both DSP and acoustic treatments can be useful under some circumstances, and with some limitations.

DSP is the more magic technology.  I find it amazing the way I can simply dial in a few notch filters to fix serious problems with my speakers that could take weeks of frustrating work to fix.  And within the range of tasks DSP can accomplish, it has amazing dept, it is amazing to be able to make things like notch filters with a Q of 10 (which seems to be about right for one semi-tone, 8.9 works for two semi-tones).

Now I have come up with a DSP answer to the old DSP limitation of not being able to flatten the frequency response everywhere in a room at the same time.  It's not a perfect answer, but it's better than nothing.  And it can't be worse, because you can simply turn it off (in my case, with a remote control).  I call it Boom Control.

Some recordings need this, others are better without it.  Typically, you don't mind the room boom outside the listening position while listening to an acoustic recording.  But you may find it intolerable while listening to a recording with electric bass.

It seems that 6dB make a nice filter depth for this sort of thing.  I have room boom outside the listening position principally in the region 40-55 Hz.  For some recordings with electric bass, it seems helpful to engage a 6dB filter with Q of 1.6 at 42 Hz.

Unfortunately, although intended to reduce the boom outside the listening position, it does also reduce the bass frequency response AT the listening position.  It may make some rock recordings sound slightly anemic.  On the other hand, it may make some otherwise intolerable recordings very listenable.  So, it's a compromise, but it works.

The effect of the reduction around the room periphery seems more than I would have expected from a 6dB reduction, while the bass at the listening position seems reduced a little but not 6dB.  Basically, you switch on the Boom Control, and all the bass rattling (or most of it, anyway) goes away.  Suddenly you have a real image with bass instruments, etc, right in front of you where it should be.  That's how it works sometimes.  Other times, you feel "where did the bass go" and then you switch off the Boom Control and you get all you want back again (because perhaps a lot of it was in the 40-55Hz range being attenuated).

Boom Control is essential for the record Bass Ecstasy (which, btw, I find quite fun to listen to in a guilty pleasure sort of way).  On Dark Side of the Moon, I prefer to leave the Boom Control off because the engineering of the record has already eliminated the boom and it sounds a bit too dry with Boom Control.

I have dialed in this filter into EQ #1 on my Tact RCS preamp.  I can turn it on or off with remote control.

Tuesday, February 8, 2011

Buzzes fixed with notch filters

Yes I know now my Acoustat speakers should eventually be taken apart, as much as can be done, and have their membranes heat treated to restore original tension.  Hopefully that will get rid of the buzzing resonances at 116Hz and around 172 Hz.

And in recent posts, I've described how I moved the cutoff frequency for the Acoustats from 104Hz up to 121Hz to get rid of the buzzing (that was good for a Tact level of 83dB out of 99.9 on "A Story Within A Story" by Pat Metheny).

That's a neat trick, but added to the midbass depression around 100Hz.

Last night I took a slightly more direct approach and used the Parametric Equalizer option on my Behringer DCX 2496 digital crossover to notch out two notable buzzy resonances in the Acoustats.  I added a -15dB notch with Q of 8.9 at 116 Hz, and a -6dB notch with Q of 7.9 at 172 Hz.

With these notches in place, I could restore the crossover point to 104Hz (restoring some of the lost midbass) and still play A Story Within A Story as a Tact level of 93dB without getting the buzz.  That's more than 10dB of added dynamic range.

With the "Sine Wave" program I created on my Kurzweil K2661 synthesizer, I can easily go up and down the keyboard to identify the problem buzzes and other resonances, and then test how well I've notched them out afterwards.  Before the notch, I did not want to press the A2 or A#2 keys for fear of getting painful sound.  Afterwards, I can play all the keys in that region, and they are all about at the same level, with A2 and A#2 only sounding slightly depressed.  I identified the need for the second notch at 172 Hz purely by using the keyboard.

The sound with all the new DSP programming I've done recently is incredibly clean, spacious, enjoyable.  The dynamic range is far greater, and I can listen to fairly heavy bass without cringing.  In addition to the Pat Metheny, I listened to Dark Side of the Moon and Bass Ecstasy.  DSOM was like hearing it for the first time, there was so much more inner spaciousness from recent changes, including more forward listening position.  Bass Ecstasy could finally be turned up loud enough to be really enjoyable.

Bottom line: if you do not fix buzzes the "old fashioned way" by eliminating them at the source, it is critical that you do something about them anyway using notch filters.  It's worth sacrificing a tiny bit of frequency response flatness to get rid of painful and potentially destructive resonances.  Little is more annoying than having your loudspeakers buzz.

(BTW, the legendary BBC-designed LS 3/5A speaker actually has an very terrible resonance in it's Bextrene woofer around 1kHz.  The KEF B110 woofer was heavily treated, but still has the resonance, so it gets notched out in the crossover, one of the reasons the LS3/5A crossover is so complex.  So many of the most highly appreciated audio products depend on the same kind of tricks that I use.)


Playing on the Kurzweil, which is in the corner of my living room, I was struck by how loud the resonances around 42 Hz are, even after full Tact room correction.  It turns out that these resonances (which span the range 35Hz to 50Hz) seem to be most offensive in the corners and elsewhere around the periphery of the room.  In the central portion of the room, these resonances are not directly audible.  The Tact system is only correcting the resonances detected from the position(s) of the microphones.  It can't correct resonances that only increase the SPL somewhere else in the room.

Actually, the Tact does have a multiple-measurement feature which I believe is intended to deal with problems like this.  But I'm not clear on how to use it or how it works yet.

A problem like this is something that requires a compromise solution of some kind.  To reduce the BOOM around the room it may be necessary to take a slight hit on the frequency response at the listening position.  But how much?  Well, that's up to me, a computer can't make the decision.

One obvious solution might be split-the-difference.  Assuming flat response at the listening position, but 20dB peak in the corner, chose a 10dB adjustment.

I tried a 6dB solution like this.  I had figured out long ago that a 42Hz notch with Q of 1.6 worked nicely on these resonances (it had already been dialed in on my DCX, but was turned off).  So I pulled that region down by 6dB.

Unfortunately, it still sounds boomy in the corners, and 6dB reduction around 42Hz in the listening position makes for a wimpy bass sound on some recordings, while still sounding stressed on certain others (like Bass Ecstasy).

A better solution here would be bass traps to reduce the resonances acoustically.  But those are expensive.

For now, I'm going to have 3 EQ levels dialed into my Tact as remote control selectable options.  A 3dB reduction (barely noticable loss of bass, but helpful reduction in boom at periphery), a 6dB reduction, and a 10dB reduction.

For the real kick-back bass, I'll turn the EQ off.  For entertaining guests, I might use the 10dB reduction.


Both of the above are uses of digital signal processing (DSP) to deal with specific problems.  Now when one is also using correction, that raises an additional issue.  Should one do the correction measurements with the notch filters in place, or just add them in afterwards?

It occurs to me that in cases like these it is desirable to add the notch filters afterwards.  Thus these are "post correction filters".

If the notch filters are added in prior to correction, the correction system may attempt to un-notch them to some degree.

However, it may be that the correction system, by design, doesn't do much for notches.  It is longstanding conventional wisdom that audio response notches should not be boosted.

Given lack of understanding about how Tact actually works, it might be desireable to try this both ways.

Likewise for the peripheral room correction.  This is a nice kind of thing to have as remote control selectable option as a post-correction filter.

In principle, the need for post correction filters is reduced by the Tact having "room curve" facilities.  You could draw needed notches right into the room curve.  But once again, not knowing exactly how to draw a Q of 8.9 notch makes this a bit tricky.  Also, it's far easier to make and test small adjustments by turning knobs on the Behringer than running the complicated RCS 2.0 software.

Nothing beats an automated system like RCS for flattening (or curving in desireable ways) the overall frequency response.  And it works OK on most resonances.  But some things require a bit more focus to do correctly, and in tricky situations setting up notches and other post-correction filters manually after full correction might be the way to go.

I would like Tact better if it were a more open system, say for example if you could chose which correction algorithms you wanted to use.  But it is still far better than "fully automated" systems in that Tact lets you can draw target room curve, and see what measurements are being made so you can fix things up prior to correction.

There has been a trend toward fully automated DSP systems recently.  I don't trust those at all.

Monday, February 7, 2011

Quad Esl-63 impulse response

Here are the Quad Esl-63 impulse tests done by John Atkinson (halfway down page):

His actual impulse response test is done with 55us pulse.  That would correspond to about 19kHz frequency, thus it does not show effects from overall group delay down to 20 Hz.  Nevertheless, what you see is a big positive pulse followed by a smallish overshoot, about 15% at most, but which lasts a bit longer then the initial impulse.  Thus from one cycle input you are getting two out, but the second is greatly reduced.  He claims that the length of the tail on the impulse response corresponds to 12kHz resonant frequency.

I wonder what this would look like with a 1ms pulse.  Probably more complex, like I get from my Acoustat.

Then he shows calculated step response.  If reproduced accurately, that wouldn't even be a half cycle.  But it does show initial decay in about 1ms, overshoot, and some LF resonance after that (drumhead?).

He shows pretty nice squarewaves also, with 12kHz ringing.


What can/should a loudspeaker impulse response look like?

  If you have a positive unidirectional pulse, which seems to be what the Tact actually uses, what do you get from a speaker through a microphone.  It cannot be an identical positive unidirectional pulse!  The speaker/microphone system is a bandpass system which has to, at least, be adding at best about one additional "cycle."  It cannot reproduce DC, so a signal with DC offset has to be bent somewhat around that limitation.

The response you could get could start with a brief negative leading edge cycle, the positive cycle of the large body of the pulse itself, and a trailing negative restoration cycle.  So 3 half cycles from 1, and I think that's about the best that can be done from any speaker without DC capability (the best actually might be more like two half cycles, depending on duration).  Assuming the pulse is long enough to have significant low frequencies, like 1-10ms, so perfect reproduction of the pulse would otherwise require DC capability.  (The Tact seems to use two pulses in measuring actually, for higher and lower frequencies, and have separate correction algorithms for each.)

In fact, that is the kind of thing I see if I measure bandpass curves electronically from my crossover, not even going through speakers.  In fact, doing this led me to believe it is not a good idea to use very high order Linkwitz Riley LR48 in the extreme treble.  You get at least an extra cycle of ringing from that, in the electronic signal itself, and Butterworth 24 gives about the cleanest impulse short of single pole.  Strangely, at lower frequencies used for my subwoofer crossover, the high order LR48 high pass impulse electronic response looked more OK, or at least just lays on top of the panel response less objectionably.  Note that you can't entirely judge a crossover by looking just at its high or low pass section in isolation; used together the combination should approximate some kind of ideal.  But the ideal ideal is most often an "all-pass" response which shifts phase, only acoustic 6dB/octave crossovers can do better.  Nevertheless, in practice the ideal will not be achieved, so it's best if each drive signal stays as simple as possible.

After much work (!), that's also what I can get from my very complicated system.  Actually the impulse looks like about 3 full cycles, with the subsequent two being greatly reduced in level, plus the usual digital aliasing stuff around the edges.  My truly great achievement was getting a combined system response, including sub and tweeter playing at uncompromised levels, that has an impulse response that looks barely different than the Acoustat alone playing by itself.  I never expected that, I only expected that the acoutstat-by-itself would look a bit simpler, because I hadn't thought about the issues very much.

And another enduring question is what is the correct polarity?  I believe the correct polarity is not that of the leading transient, which may be always out-of-polarity in the ideal bandpass case, but that of the larger square wave what follows.  BTW, I inverted the tweeters to give them a puzzle-fitting response.  Just by themselves, the acoustats seem to have out-of-polarity leading edge and half cycle followed by full positive cycle, somehow I needed to invert the tweeters to make it fit precisely.*  (Strangely, the tweeters actually have a kind of quasi-DC response... More about that discovery later...)

I wonder if some would claim the correct polarity has the leading edge in polarity, so I could be wrong.  Perhaps it depends on the number of octaves in the bandpass, and what the bandpass functions are. And perhaps also it becomes fundamentally a subjective question in the case of a system with sufficient group delay (as probably most are).   Do you want the bass in polarity or the treble in polarity (assuming you have to make the choice)?  If the leading edge defines the treble, but a larger partial cycle follows in opposite but correct polarity, that indicates the treble is out-of-polarity but the midrange--if-not-the-bass--is in polarity.

Another interesting (and now very important) issue is the fundamental panel resonance.  I believe there is some fundamental panel resonance in the region of 55 Hz.  That resonance is how the speaker maintains frequency response to 40 Hz in spite of being a narrow dipole, I think.  Well, it has that resonance when manufactured, but age probably causes membrane to get loose, hence less controlled, which may mean higher resonance harmonics, and that is what seems to be happening, I get buzzing around 110 Hz (on one recording anyway).  Well, I need to fix the speaker, the Acoustats use HS65 which is actually 6.5 mil heat shrink.  The fix is to use hair dryer to shrink the membrane back to tightness.  It's supposed to last about forever if you keep doing that.  Other than that, the panels cannot be repaired, you can scavenge old units or make new ones using the same principles (people have claimed to do that and say they never want to go back, and one beauty of the acoustat design is that there is nothing in it that couldn't be done in a garage with readily available and cheap materials).

In the meantime, and this is what I did before and now discovered I must do, I raise the crossover point to 121 Hz and the problem disappears at reasonable levels.  I thought I could get away with lowering the crossover point (and with brand new panels, I ought to be able to do so) but apparently not for now.

But anyway, my original question relates to this because we have to consider the panel not as a DC tracking system but a fundamentally resonant system, with fundamental low frequency "drumhead" resonance and high frequency "breakup" resonance.  These affect how impulse is going to look, even in the absense of crossovers, reflections, diffractions, etc.

Speaking of which, I should dig out the old impulse picture from an ESL63, which I thought looked pretty good.

Room Correction Weekend ends with A Story

I did some other fun things also, but the bulk of the first weekend in February 2011 was spent making microphone measurements of my audio system using my Tact Room Correction System (RCS) 2.0 Preamp, making correction curves, and listening to them (Saturday night didn't end until 7am because I couldn't quit listening because it sounded so good).  A gazillion measurements were made, and I made many new important discoveries.   Photos were taken of many graphs, but it will take a week just to sort through them.  It ended on a mixed note, however, so-to-speak.  Playing through Sonos the Pat Metheny track "A Story Within A Story" revealed a buzzy bass note in the intro.  Damn.  I need, someday which will probably not be soon, to take the Acoustats apart and give their membranes the hair dryer shrink treatment.  However, I can fix the buzz by running the Acoustats as I did last year, Crossed over above 120hz.  (Actually, I think I previously crossed at 116, but with current correction curves I need to cross at 121, and higher might be even better wrt reducing potential for buzz.)

This time, since January I had tried to cross the Acoustats in at 104Hz to avoid a disturbing room resonance.  The subs cross out at 85Hz also to avoid that resonance, the resonance pick up the tab in the middle giving reasonably smooth response (and nicely boomless) without correction.

Great idea I thought, which could be applied to most speaker systems: stagger the sub crossover around the room ceiling resonance around 100Hz.  And the Acoustats are a "full range" speaker (as some people define it, anyway) that has audible bass (with room gain) down to 40Hz or so.  So there not only shouldn't have been a problem changing crossover from 116 down to 104Hz, that should give the system cleaner "panel bass" (actually, many people think panel bass is fake sounding, but flat speaker lovers usually think cone bass is fake sounding).

But there is a problem, because of my 20 year old panels and my desire to listen at pretty substantial (not ear damaging) levels, and because of room correction itself.  Because the Acoustats have a deep midbass depression in their response between 110 and 400 Hz, the room correction is fixing that with midbass boost.  That midbass boost is pusing the speaker into noisy distortion around 110Hz.  Without the boost, I wouldn't get the distortion without playing considerably louder (though I have not tried this, I simply switched to correction Bypass and the buzz went away at the same level).

But the panels clearly have a problem, and even if boost is required to make them sound bad on A Story Within A Story, they could be distorting less noticeably on other music, and in some cases, without boost.

You could pin the problem on the crossover point, the room correction, the loudness level, and ultimately the speakers.  In the sense that the speakers shouldn't have this problem at this frequency and level, it is a speaker problem that ultimately needs to be addressed (and yes I have even thought of buying a totally different kind of speaker, like Magnepan 1.7's or Linkwitz Orion).  But meanwhile, I can work around it by judicious choice of crossover, and as long as I don't play too loud.  Using current room correction curve (measured for 104Hz panel crossover) #2, but with post-correction change to 121Hz crossover, I can play A Story Within A Story to 81 gain level on Tact preamp without distortion.  At 82 the distortion is barely audible, at 83 he distortion sounds like "just a normal part of slap bass" (but it's being exaggerated by speaker distortion to sound qualitatively different than the actual recording).

I don't want to give up the correction.  The perfectly calibrated midbass boost brings life back to music.  The music sounds so much more real with good midbass, even if I have to compromise that midbass slightly by leaving a hole in the midbass between 104Hz and 121Hz by moving the panel crossover up to 121Hz.  That is peanuts compared with having the whole range depressed.  By the way, I think the whole range depression may result partly from the infamous dipolar cancellation.  Linkwitz deals with this in his design by calculating the effect and deliberately equalizing it.  I'm canceling the effect by measuring the system and room correcting it.

So I'm listening with the small hole in the midbass instead of the big midbass depression that I had previously.  Eventually I'll have to go back and do a whole new series of corrections based on the new crossover.  I was going to tell you how much work that was this time (I did two sets of 6 corrections over the weekend...I had to run a second set because I made a couple of mistakes in the first set; each set has two measurements (for helpful redundancy) for regular, no supertweet, and no sub or supertweet conditions).  But I before I do that, I also need to see if I can push the subwoofer to cross over slightly higher.  That work will require more measurements and tests too.

I can just press the Bypass button on the Tact remote to bring uncorrected response.  It's no longer the uncorrected response from January, as prelude to Tact correction I changed tweeter highpass to 20kHz and changed from LR48 to BU24, level reduced by 2dB net, and fine-tuned crossover delay.  That made for something like perfect impulse reproduction, at least with the supertweeter adding to the Acoustat nicely and even making impulse sharper.  And now, of course, bypass now has both the hole in the 104-121Hz midbass AND the 110-400Hz depression, whereas before it just had the depression, so now bypass is slightly worse in the bass, arguably better in the treble than it was before.

Anyway, though the bypass is slighly different than before, I don't believe it has gotten much worse.  But compared to the corrected response, you just don't want to listen to it anymore.  Sure it's very open.  But it's very thin and bright sounding, with too much highs above 1kHz (just consistently bright) and no midbass.

That was why I do room correction!

While great progress was made, now I know that even more work will be required that I was expecting to be sufficient, with no end in sight (such as working on the speaker itself).

But that's the way life is, isn't it?  If there were nothing more to be done, how fun would that be?

Saturday, February 5, 2011

OK, perhaps I should worry about group delay

After studying the loudspeaker crossover issues in the 1977-1983 time frame, I came to the conclusion that the Linkwitz-Riley was by far the best kind of crossover for most drivers.  Generally speaking the 24-dB per octave (LR24) would be the best crossover choice.  Linkwitz concluded that the group delay (a compromise) introduced in the overall response of the crossover (by design, the drivers are always in phase with each other through the crossover region) was not audible and of negligible importance.

Only a few crossover designs, like the 6db per octave acoustic crossover used in Thiel, Vandersteen, and some others, can achieve total  lack of group delay in the summed response.  And that requires multiple other design compromises.

When the Behringer DCX 2496 digital crossover came out, we were treated to LR48 achieved in hirez digital.  What could be better?  That's what i have adopted uncritically since 2005; I now use the Behringer in both living room and bed room systems.

Now, I have figured out how to examine the impulse and frequency response of the crossover network by itself, and I am not so sure LR48 is the best choice.  For augentation purposes, where there is no perfect cancellation of phase artifacts, Butterworth 24db per octoave (BU24) looks like the choice that gives the best compromise between steep cutoff and lack of visible time dispersion.

LR24 might be a better choice if you had a perfect acoustic LR24, something very hard to achieve in practice,  but given lack of perfection, the best bet is probably to minimize time dispersion in each crossover member with one that provides less dispersion

In a quasi-augmentation mode, or as a solo highpass network, the LR24 increases time dispersion AND reduces sharpness of cutoff compared to BU24.  Im not sure of the advantages of the Bessel, it may have he steepest cutoff, but it is marred by 20dB passband irregularities.  It might be OK for supertweeter where, say, above 20K you don't care about passband irregularities.

I think it's possible that in the low frequency crossover (the highpass on the acoustats is currently set to 104Hz) LR48 works OK, and is extremely beneficial in reducing panel flap at high volume levels.

The visible difference between BU24 and LR48 is pretty small at 104Hz, the LR48 does have a bit of initial out-of=polarity undershoot, and somewhat more overshoot on the trailing edge followed by slow recovery.  Whereas the BU24 almost looks untouched, the perfect Tact impulse (that is, as perfect as the Tact gets) almost. But it just doesn't phase me much, at least on screen.

But what has been driving me apoplectic about this is the supertweeter highpass.  It turns out that the highpass signal from LR48 at 15.5 or 20 kHz has 3-4 cycles of ringing at 20kHz.  That's all there is, it doesn't look like an impulse at all, just ringing.  Superimposed on a perfect impulse, it smears it out considerably.  My time domain purist friends should be laughing at me now.

That's exactly what I've been seeing in the system impuse time response.  And it bugged me so much I refused to print it yesterday.  Now I've figured it out, at least partly.  Even if the supertweeter is reproducing the signal that it receives perfectly, that impulse response looks like 4 cycles of ringing at 20kHz because that is the signal that the crossover is providing.

Now in the context of a perfect LR48 crossover, with perfect high and low frequency drivers crossing over, the majority of the phase anomalies might well cancel out (I am not entirely sure of this...) and it wouldn't look so bad.  Maybe.

But in the context of the kind of slap dash (if infinitely pondered) systems which are the only kind I can put together, not being able to hire a team of engineers, it's looking to me like simpler is better, probably BU24, the time domain smearing is cut in half or less, in fact it doesn't look like ringing anymore, it looks like a double pulse, which probably adds nicely with the low frequency system pulse.