[Springbok10]

Hi,
I have been posting (endlessly, it seems) on A'gon in an attempt to find a speaker to fill my room, 23X17X15 ft ceiling, to replace my Kharma 3.2 FEs (just sold). The Kharmas did not fill the room and I am sure were not designed to do so. I am obviously now looking at the Revs and have spoken to Bob and Steve. My minor concern is placement - they cannot be more than 6-12" from front wall, but my major concern is that I have anAtma-Sphere MP-1 preamp and Atma-Sphere MA 2.2 220 WPC OTL amp. Will this amp fulfill the obviously vast potential ofthe Revs? Obviously, the OTLs perform best with 8 ohm or higher loads. Please give me your experience of using tube amps, especially OTLS, with the Revs. Also, will only 6-12" clearnce ruin the soundstage? (Speakers will be 8 ft apart and listening position at 11 ft).
Thanks!

[ted_b]

Hi.  Welcome to AC.  We've communicated earlier via email, as you well know.  Although I have no experience with the Atma's I know they are great tube amps.

My reason for posting, though, is to continue this dialog:  what are the reasons you have to stay so close to the front and side walls, yet such a nice big room?  I'm slightly concerned that the Revs would be a bit crowded having to hug the perimeter so closely, and that maybe the vast SP Tech knowledge here would have some input.  Maybe another SP Tech speaker does better with that restriction?  Dunno. 

[Karsten]

I have been running my Revelations with both Joule Electra VZN100 and VZN160 with a pretty good result. No real problem in driving them at all. In absolute terms there is of course a trade off in the bass control using OTL's but this is hardly the speakers fault.

Going too close to the rear wall will have an impact on the soundstaging, but similar effect with all other speakers which has been tested here. The general rule is to take the length of the room and divide by 7 or 9, this is usually the approx. position where you get the best coupling to the room.

Regards,
Karsten

[JP78]

you may want to talk to duke of audiokinesis...he does some wonderful things building a small circuit to correct the atmasphere output for his own speakers.  being the gentleman he is, i'm sure something could be arranged for those beautiful monoblocks you have to match the revelations.

better yet, duke could send the schematic over to bob, who could then incorporate this into the revelation external crossover box.  how great would that be?!

[Springbok10]

The room placement restriction is because of furniture size and placement (wife issue) - when I allowed her to purchase  new, heavy, high, upholstered furnture I did not realise that it would ruin my sound! The room placement is of course an issue with nearly all speakers, and although I was hoping to get by with 6-12", it may negate the positives of these speakers. The space is further hampered by bookshelves on both ends and a huge wooden box, 7 ft long, hinged, used as a bench seat when amps not in use, to hold the monoblocs. The speakers would be at both ends, about 2 ft from the side walls.

[ted_b]

Could you draw a picture or take a few snapshots?  We'd LOVE to convince you to spend $$ on Revs, but the nice thing about the SP Tech line is that you get the same sound, only smaller (just kidding; mostly it's an lf extension thing) as you move down from the Revs.  To put a Ferrari in your stable only to have it go 40 mph is not a good thing....but it IS a Ferrari, nonetheless.   

[Springbok10]

I would be happy to post pictures - when I am back home on Monday - I'm travelling now - how do I post them? I suppose I will have an issue with any speaker except maybe the old Allison Ones that I threw away 20 years ago..........
The laws of Physics are a real problem. Even Acoustic treatments won't give me a deep, wide soundstage nor restore the midrange clarity of those confining side-walls. Right, Bob?

[Duke]

Let me clarify a bit about the circuit referrred to by JP78:  It wasn't for my speakers; it was to smooth out an impedance hump in the midrange on speakers owned by a friend of mine, so that they would be more compatible with the Atma-Sphere S-30 (which is much more speaker-impedance-sensitive than the MA-2 is). 

I'm under the impression that SP Tech speakers have exceptionally smooth impedance curves, and if so I'd expect the Revelation system to work very well with the big Atma-Sphere amps.

If (by chance) the Revelation has a one-hump impedance curve (ignoring bass impedance peaks for now), then I could probably conjur up a circuit to smooth it out.  If it has more than one hump, I couldn't do the design from a distance.   I don't have a circuit that addresses impedance dips.  But my impression is that Bob Smith's impedance curves are smooth enough that this won't even begin to be an issue. 

Now it's possible that the Atma-Sphere amps might result in elevated bass, if there are significant impedance peaks in the bass region (room acoustics could also boost the bass).  The Atma-Spheres will interact differently with these impedance peaks than a solid state amp would.   I'm not familiar enough with the bass systems that Bob uses to make a precise suggestion should some adjustment be needed, but in a transmission line system we'd probably want to increase the stuffing density and in a ported speaker we'd probably want to lower the tuning frequency.   

As far as soundstage goes, imho a controlled-pattern speaker like SP Tech will outperform a more conventional design when positioning is less than optimum., as in this case.

One final comment - I looked at the specs on the Revelation, and saw there a level of honesty that is quite unusual:  Impedance is 4 ohms, efficiency is 88 dB/1 watt, and sensitivity is 91 dB/2.83 volts.   Most makers of 4-ohm speakers only give the voltage sensitivity because it is higher, and many misleadingly call it "efficiency".   Bob has given you the whole picture.   The MA-2 will do 220 watts into a 4-ohm load, which should give you unclipped peaks in the 111 dB ballpark.   

I'd want to double-check with Bob on the impedance curve just to be sure before committing, but the MA-2 is pretty forgiving in this area.  Assuming the impedance curve looks good, I think you'll have a truly world-class combination there.  You'll have superbly natural and relaxing timbre, rich inner detail, and incredible dynamics.   

Duke

[billc]

If you are still considering options for high power tube amps, I highly recommend the Antique Sound Lab Hurricanes (I or II version), used in pentode mode (200w/monoblock), especially in used with SP Tech speakers.  The SP speakers like the higher damping provided in pentode mode.  And when modded by Response Audio they are amazing; I offer my perspective as an owner.  With mods, they are astonishing.

But, the previous comments are very insightful.  Placing any speaker driven by high power that close to the front and side walls is asking for acoustic problems.  I have a 12x24 room, and have SP Mini's about 3.5 feet from the front and side walls; this placement is important to their quality sound.  And even though they are the Mini's, they definitely fill the room.  (Although the room is fairly reflective.)

It may be other issues that affect your goal of room-filling sound, not simply speaker dimension, or amp wattage.  The Rev's are great speakers, to be sure, but some careful consideration should be given to room/speaker/amp interaction.  The Rev's are a good place to start if you are seeking top shelf sound, but positioning will be important to optimal performance.

Bill C

[chadh]

Perhaps it would be worth thinking about the Speltz zero autoformers to raise the impedance that you amps will see.

Chad

[Aether Audio]

Springbok10,

Let me chime in here to clarify some things.  First of all… thanks Duke for your input!  Duke is a very knowledgeable guy and he’s pretty right on about the placement issue and all.  Before I get very deep into the subject (you’ll learn to be more careful next time – as the rest of the guys can testify to ), let me just say that as Duke pointed out, our Revelations or any of our speakers are probably one of the least sensitive designs available with regards to room placement.  In other words, if you don’t have an ideal placement situation, you REALLY ought to consider our stuff above a lot of the rest.

In your situation, two of the WORST designs that you would definitely want to avoid would be either an omni (as in say…mbl) or a dipole (planer electrostatic) or even an open baffle design.  There’s really only one design that would stand a decent chance of sounding good and that would be your standard “monopolar” speaker.  Now, if you could find one of those that also offered some directivity control, then you’d have a real good chance of achieving quality sound.  Hey…whaddaya know… that’s just what our designs offer!!!

OK, down to some math.  Our Revelations are 15.75-inches deep and 12.5-inches wide.  Divide the width by 2, add the depth and then add the 12-inches you say that you are able to pull the speaker away from the back wall.  By my math that adds up to 34-inches.  Divide 13,618 (the speed of sound in inches per second) by 2 x 34 to get the frequency that is equivalent to a 68-inch wavelength.  That works out to almost exactly 200Hz.

So…we find that we have a reflection path length of ½ wavelength at 100Hz from the front-center of the enclosure to the wall behind the speaker.  Now, if 100% of the reflected energy were to return to the point of origin (the woofers), the system would exhibit a perfect null (a complete loss of SPL) at 100Hz.  But… we must remember that the wave front is expanding as it leaves the woofers as well as the entire time it is traveling to the back wall and then reflecting back to the woofer again.  That means that far less than 100% of the reflected energy will likely return to the drivers.  If we assume that as much as 50% were to return (somewhat unlikely) then our system would exhibit a 6dB reduction of SPL at 100 Hz.  A more likely amount of 25% would yield a 3dB reduction of output.

This all adds up to the fact that the odds are a pair of Revelations placed 1-foot from the wall behind them would exhibit a dip of only about 3 to 6dB at 100Hz.  This is a worst-case scenario, as any phase nulling will decrease in magnitude as the frequency being reproduced is lowered from 100Hz.  As the frequency is raised from 100Hz the radiation field will change from 4pi (omni-directional) to 2pi (hemispherical) as a result of baffle width and then on to less than 2pi as a result of waveguide control.  We might get a bump in output at 200Hz as well, but again it's magnitude shouldn't be much more than about 3dB or so.

To be sure, the Revs would benefit by being pulled out further from the wall, as this would push the first reflection null down to a lower frequency.  Although...and this is something a lot of you guys probably haven't thought of... pulling the speakers out away from the wall will lower the first reflection null frequency - it is also likely to make it WORSE! This is because more of the reflected energy is likely to converge with the speakers if they are farther out in the room.  Try using an SPL meter and some test tones if you don't believe me.

Regardless, in your situation Springbok10, (specifically BECAUSE the speakers will be closer to the wall behind them) the magnitude of any null will likely be relatively mild and well within the frequency “averaging” nature of human hearing.  100Hz is also at the very bottom of the midrange and therefore our hearing is also far less sensitive to any frequency variations in that range.  Actually, a small dip there might be a good thing as a lot of rooms often have a peak there anyway.

The other consideration is the consequence of early reflections affecting/contaminating the time-dependant spatial (imaging) information in the recordings being reproduced.  This is always a concern to some degree, but it is important to note that it has been found that human hearing has a very great ability to separate and identify the first arrival signal from those generated by reflections.  This is how we are able to locate the origin of a sound source in a darkened room.  Were it not for this ability, we would be at a total loss in trying to identify the location of a friend’s voice in a crowded, noisy space.  Again, due to the directivity control afforded by the waveguide in our designs, they will exhibit far less sensitivity to this problem than virtually any other speaker design.

The upshot is that the worst artifact to result from such placement would be a boost of the frequency range below the 100Hz null – i.e., the bass range (100Hz on down).  Well… that’s pretty easy to compensate for.  In fact, I can easily design and build a passive network to place in line with your power amps (that would be a passive, line-level equalizing network).  Beyond that, if I can find out a few specifications regarding the output impedance of your amps, I can run some simulations in SPICE (circuit modeling software) to find whatever compensation networks would be needed for you amp.  Be it impedance compensation for the speakers or frequency response correction for first reflection boundary gain – we can fix it - piece of cake.

The bottom line is… don’t let either the amp or room situation scare you.  We can tweak the Revs to as close to perfection as any speaker could ever be. 

Now… will somebody please give me a REAL challenge!  Nevermind.  Already got one.  It’s called MAKING MONEY!!!

-Bob

[ted_b]

See Springbok10, I told you he was a mad scientist genius.   

[Springbok10]

Yes, I see.............

[Duke]

Bob, I appreciate your very thorough and professional post.  I really like the way you analyzed the effect of the reflection off the wall there - I had neither done the math nor taken into account the expanding wavefront.  So I'm taking notes...

Anyway, the output impedance of the Atma-Sphere MA-2 amplifier is about 1.75 ohms.   The fact that you'd even ask shows how thorough you are. 

Given the speakers' proximity to the side and "front" walls in Springbok10's anticipated setup, would you be inclined to go for a lot of toe-in, or not?  Of course it's easy to try both ways and see which works best, but I'm curious about your thinking on the subject.  As you probably saw in Denver, I like toe-in.  But then maybe I'm like the guy whose only tool is a hammer, so to me every problem looks like a nail.

Duke

[Karsten]

Springbok,

I may ad that one setup which works really well with the speakers very close to the rear wall(s), is a diagonal setup. I don't know if it is a possibility in your room though.

Karsten

[ted_b]

I like toe-in, too.  My Revs are toed in substantially and I get great imaging from it.  The sweetspot might be a tad narrower, but I don't care. 

[Springbok10]

Thank you all for your input; Bob, your analysis is amazing, and it really says a lot about both you and Duke, paricularly since he is a competitor, that you would both be so detailed and forthright in your discussion about my needs - I am very grateful and quite frankly, amzed, since I've never seen 2 competitors have such a frank discussion publicly before. Karsten, yes, I can place them diagonally - but where would the listening position move to? Currently I have 8 ft between speakers and listen at 11 ft.
Bob, please explain again how we perceive soundstage from your analysis and why it's better further out? Somebody on A'gon suggested Tetra 606s, which are 8 ohm, but they are dipoles and should be worse close to walls, or not?
I'll be home tomorrow and will decide soon and put you all out of your misery:)

[Duke]

Speaker manufacturers are like fishermen all fishing in the same pond, and imho they might as well get along with one another.  Just so happens Bob and I are using very similar bait!  I think we both see that as an excuse for kinship and even friendship. 

Springbok10, my understanding is that sound source localization is mainly determined by the first .68 milliseconds (9 inches) of a first-arrival sound, after which the "precedence effect" kicks in, largely suppressing directional cues from repetitions of that original sound (i.e. reflections).  But early-arrival reflections still can have an effect; reflections arriving within 10 milliseconds of the direct sound (path length of about 11 feet) tend to be detrimental to imaging and timbre.  The direction of the reflection plays a role in whether it effects imaging or timbre the most.  Later-arriving reflections tend to impart a sense of richness and spaciousness with little or no detriment (assuming their spectral balance is in the ballpark - waveguides shine in this area).   So imho what we'd like to do is minimize reflections arriving earlier than 10 milliseconds behind the direct sound, but still have plenty of late-arriving reverberant energy.

One way to achieve this is to use a loudspeaker that has good directional control which will tend to minimize early reflections, and then have a fairly lively (and preferably diffuse) room so that the reverberant energy decays rather slowly.  Waveguides are imho the most practical way to get good directional control with low coloration.  Still, I think that energy bending around the side and bouncing off the wall behind and between the speakers is detrimental to depth of image if it arrives too early.

Dipole and bipole and omnidirectional speakers would not be my choice for a situation like yours because their extra off-axis energy would begin to arrive at the listening position too early.  Having sufficient distance behind such speakers makes a big difference.  If you notice how MBL shows their big omnispeakers, it's typically in a large room with plenty of path length distance to the nearest walls. 

Duke

[Aether Audio]

Duke,

Thanks my friend.  Looks like I definitely owe you on this one. Assuming the amp’s output impedance is fairly linear – at least across the midrange – I found that after plugging 1.75-Ohms into SPICE, the amp/Revelation combo ended up exhibiting a significant impedance peak at 2kHz.  That peak would have resulted in a bump at 2kHz of between 1 and 2dB.  Well, that wouldn’t do so I developed a network that would flatten the impedance rise.  Looks like 500uH, 20uF & 13-Ohms will just about do the trick.

Actually, it turned out that the network helped keep things in the 7 to 9-Ohm region across the entire band above crossover (700Hz).  In fact, the model shows that the impedance only drops below 6-Ohms down below 100Hz.  Although… the model doesn’t include the reflected acoustic impedance at low frequencies, but that will depend on room placement to a certain degree anyway. 

So… it looks like we have a solution for the Atmosphere amps!  In fact, we can do the same for any amp – if need be.  Most solid-state amps won’t require it and/or benefit, as their output impedance is typically 0.1-Ohm or less.  But… any amp that exhibits an output impedance of 1-Ohm or greater might sound a little smoother through the midrange with an impedance correction network.

Springbok10,

Well… there you go. That problem is solved. We’ll build the network and ship it with the speakers - if you want.  As far as the Tetras are concerned, you are absolutely correct.  You don’t want to radiate the wall behind the speaker any more than absolutely necessary.  In my way of thinking, that means in ALL cases.  But then again, I’m a devout advocate of “Live End – Dead End” theory.  That’s what most of the big recording studios employ when they record and mix the music you listen to… don’t ya know?  The idea is to avoid all early reflections from the area around the speakers (the “dead end”) and then use a fair amount of reflection/diffusion from the end of the room behind the listening position (the “live end”).

As Duke has pointed out, early reflections are bad.  They mix with the direct (un-reflected) signal and merge into a resulting “sound” that the brain falsely thinks is the total “character” of the original sound.  In other words, if reflected signals arrive at your ear that are 40 milliseconds or less (sound travels about 1-foot per millisecond) behind the direct signal in time, the brain doesn’t interpret those reflections as “echoes” at all.  That leads to a highly “colored” sound with all kinds of dips and peaks in the frequency response.  You can read all about the “Hass Effect” here: http://en.wikipedia.org/wiki/Haas_effect

It’s bad enough that it happens to virtually every speaker at low frequencies, but fortunately our brains don’t pay much attention to it down there.  Where the effect is really bad is when it happens in the mid and high frequencies.  That’s what you get too when you use a dipole or open baffle speaker.  You end up “spraying the room” with a lot of upper band energy that comes back later in time and mixes with the direct signal.  They then sum together in ways that are so complex that it would take a super computer to model.  The sonic effect is that a lot if not most of the spatial information that is embedded in the recording is covered up or “masked” by the acoustics of your listening room. 

Question:  Do you want the performers in your living room or do you want to close your eyes and be transported to the concert hall that the musicians were recorded in?  If you want a “band in a box” then get the Tetras.  If you want to go to Carnegie Hall without ever leaving your listening chair, then get the Revs or something as close to them as you can find.  No matter what, even if you were of the dipole or open baffle persuasion, you wouldn’t want to use that type of speaker with your placement limitation – that’s for sure.

You asked about toe-in as well.  Seeing that our mid to high frequency dispersion is already being controlled by our waveguide, small amounts of toe-in have a much larger effect than produced by most speakers.  With ours, a little toe-in reduces sidewall reflections by quite a bit.  Not necessarily so with other, more common designs.  Something like the Kharmas would require a lot of toe-in to have much of any effect and/or improvement.  Our waveguide allows you to use just enough toe-in to significantly reduce sidewall reflections while avoiding a large collapse of the soundstage.  Just ask any of our owners – they’ll all tell you the same thing.

Anyway, hope this helps.  Oh…as far as Duke and I go, we have the greatest respect for one another.  Back when SP Tech first started Duke immediately saw the advantage to our designs.  So much so that a little later on he teamed up for a while with one of the worlds greatest waveguide genius’ – Earl Geddes -  to come up with a design for himself.  The way I see it, Duke earned my respect for appreciating my designs and I earned his for pointing the “WAY.”  Besides, Duke is one of the kindest and most humble men you’ll ever meet… and that’s all it takes to earn my respect to begin with.  The world is full of narcissistic geniuses but men of true character seem to be well on their way to extinction.

-Bob

[Frank S.]

I love reading your posts Bob.