IAR Hotline! 76-79:  Master Guide to the Best of 1998:  Loudspeakers Samadhi Ichiban  We now move into the realm of small speaker systems. The Ichiban, though floor standing, is downright petite, 9" by 9" by 33", a  mini-tower instead of a midi-tower. It is a very capable full range $2495 ($1,990/pr introductory offer) speaker, which can give great musical satisfaction. It has good transparency and low colorations, and neutral tonal  balance throughout the range it covers. More importantly, within these capabilities it does not commit salient, active errors that would detract from musical enjoyment. We have already seen how larger systems with greater  nominal capabilities nevertheless commit active errors or have colorations that intrude upon music. The most remarkable sonic quality of the Ichiban is its big sound, re-creating a large stereo stage filled with music. The specs for the Ichiban are typical of a small mini-tower; with a bass response -3dB point at 38 Hz, you won't get room shaking deep bass as with the larger systems above. So what gives the Ichiban its big sound, and what earns it this place in the rankings among the big guys? Dick Olsher, consulting design engineer for Samadhi, has come up with a neat design variation on an old tune. Stu Hegeman made an omnidirectional mini-tower by aiming a full range driver up at the ceiling. Dick Shahinian makes  mini-towers with a forward facing woofer and an omnidirectional bank of tweeters aimed upward. Dick Olsher has now evolved this design approach by reversing the Shahinian design concept. In these Olsher designs, the tweeter  faces forward toward the listener, but the woofer/midrange is aimed upward at the ceiling. All these design approaches create a big, spacious sound, as well as a broad stereo image rich in ambience, by providing a higher ratio of reflected and delayed (ideally beyond the Haas window) sound to direct sound (the Bose  901 aimed for a similar goal). However, prior to Olsher's twist, these design approaches also had some sonic problems. Stereo localization of specific instruments within the richly ambient stage became amorphous or imprecise.  Tonal balance of the system was heavily dependent on the reflective characteristics of the wall and ceiling surfaces in each different room, as well as room placement. Articulation of upper frequencies suffered, and the upper  treble would sound dull, since no one tweeter was on-axis with the listener. By reversing the roles of the woofer/midrange and tweeter, Olsher has alleviated these problems. In Olsher's design, the on-axis tweeter gives the listener precise articulation and extended upper treble response, as in a  standard speaker system design. The ear/brain is most sensitive to localization cues in the upper midrange, which are provided by the on-axis tweeter in the form of coherent direct sound, not incoherently diffuse reflected  sound, so stereo localization can be as precise as in a standard speaker system. Meanwhile, in Olsher's design the upward pointing woofer/midrange driver creates a large, ambience-rich sound for the lower frequencies, and also makes Olsher's system sound like a much larger speaker system. It's worth  remembering that all speakers with forward firing woofers are still omnidirectional for all bass frequencies anyway. But the midranges of these standard speakers are directional, aiming only at the listener, so these speakers  exhibit a discontinuity in sonic quality, and a discontinuity in the radiated reverberant field in your room, going from the bass to the midrange. The Olsher design keeps the midrange consistent with the bass, both being  omnidirectional from the same driver. Furthermore, the tweeters of most standard systems are nearly omnidirectional for their lower frequencies (the upper midrange); thus in standard speakers there is a double discontinuity, from the omnidirectional bass to a  directional midrange, and then to an omnidirectional upper midrange. However, in Olsher's design the midrange is kept omnidirectional by the upward firing bass/midrange driver, so there is good continuity and consistency from  the bass, through the midrange, through the upper midrange from the tweeter—all being substantially omnidirectional. This makes the sonic character of the speaker more consistent, throughout more of the musical spectrum. Samadhi mentions, in support of this speaker design approach, the fact that most live musical instruments (and vocalists too) are omnidirectional radiators at lower frequencies but become directional at upper frequencies, and  this pattern is mimicked by their speaker's design. On the face of it, this might seem a specious argument; after all, you don't want a double dose of a particular coloration in order to accurately reproduce the music signal  captured by the microphones (remember the speaker years ago with a wooden diaphragm, which was claimed to better reproduce the sound of violin bodies). But, if we think deeper and consider the role that listening room reverberation plays in reproducing today's imperfect recordings, Samadhi's pitch does make sense. If today's recordings were made in four channel, and if they  included a realistic amount of hall reverb, then we could play them back most ideally in an anechoic chamber (through four or more speakers). But today's recordings are mostly two channel. And no recordings in the foreseeable  future will include a realistic amount of reverb from the recording hall venue. That's because mikes hear differently than we humans do; they cannot analytically separate frontal direct sound from side reverb sound as we can when listening to live music from a typical seat in the middle of the hall, where  the true sound of live music we are hearing is actually mostly reverberant hall sound. A mike placed at such a typical mid-hall listening seat would produce a recording that sounds hopelessly muddy and dull, awash with  confusing reverb. We expect from a recording the clarity, immediacy, and sparkle that we hear from the direct sound when attending a concert five. But we are able to hear this clarity, immediacy, and sparkle from our distant mid-hall seat only  because our ear/brain is able to analytically separate, in real time (on the fly), the minority direct sound (which does have this clarity, immediacy, and sparkle) from the majority reverberant sound (which heard by itself  would sound confused, muddy, and dull). Since a microphone (sensing only pressure or velocity, but not both) cannot analytically separate frontal direct sound from side reverberant sound, the electrical signal it produces  represents a conflation of the two, intertwined so they can no longer be electrically separated. If the mike were placed where we usually sit, the vast majority of this conflated signal would represent reverberant hall sound,  which would swamp the contribution from the direct sound of the music, making the resulting composite signal sound confused, muddy, and dull. Since we are stuck with today's microphones' inability to separate frontal direct sound from side reverberant sound, the only viable solution is to move the recording microphone much closer to the musicians than we would  normally sit. This increases the proportion of direct sound to reverberant hall sound in that conflated, intertwined composite we're stuck with from today's mikes. Only this can give us from a recording the clarity, immediacy,  and sparkle that we expect, that we remember being able to hear live (thanks to our ear/brain's ability to separate the frontal direct from the side reverberant sound). But moving the mikes closer creates two new problems. First, the sound becomes too dry in ambience, without enough reverberant information about the hall space and acoustic. The sonic result is that the recording portrays a  musical performance spread across a shallow stage, but not taking place on a deep stage and in a large hall. This is a consequence of moving the mikes closer, precisely to reduce the amount of reverberant sound captured in the  recording. With today's mikes, we can't get enough clarity, immediacy, and sparkle in a recording until we move the mikes so close that the reverberant hall sound is reduced to an unrealistically miniscule proportion, giving a  unrealistically dry sound. Some recording engineers try to patch up this situation with a band-aid approach of electrically adding in hall reverb picked up by remote hall ambience mikes (Keith Johnson does this trick well), but  this is not a purist correct answer, and it actually still does not add back enough reverb to produce a recording that would sound realistic without further help from further trickery in the playback system. The second problem is that moving the mikes in close picks up a different kind of direct sound from the musical instruments (and vocalists). For example, recording mikes for a symphony orchestra are typically placed above and  just in front of the string sections. This position picks up the brightest lobe radiated by the bowed strings, and proportionally minimizes the contribution of their warm wooden sounding boards. However, you almost never hear  live strings sounding this way, because there are no audience seats up in that catbird seat location where the recording mikes are. When you sit down on the main floor (the orchestra section) at a live concert, the version of acoustic signal you hear direct comes from well off the upward facing axis of each string instrument. You hear a much duller (more  resinous, less steely) sound from the strings themselves by listening off axis, and you hear a much higher proportion of the warm wooden sound from the cavity and sounding board of the instrument's body. If you sit up in the  balconies instead of down in the orchestra, you'll hear yet another tonal balance and sound from the strings. It will be more silvery, with less warm wood. But the silvery brightness will have been softened and dulled by its  passage through the great distance of air between the stage and your distant balcony seat, so it still won't sound anything like the hard, steely brightness that actually exists in the acoustic signal, up at the microphones'  close and high recording position, where no audience seat exists. Incidentally, we verified this firsthand by wandering around the whole Academy of Music concert hall while the Philadelphia Orchestra rehearsed. There are some balcony box seats on the side of this hall, well forward and almost  over the stage apron, and thus not far from where a recording mike would hang. From this sonic vantage point, the live strings sounded very different from anywhere else in the hall, and indeed sounded much like a typical  recording—artificially hard, bright, steely, and ugly—not at all like the "real" sound of live strings heard from all the other usual seats in the hall. But of course they were five and real, not an artificial  recording (so much for five music having an absolute sound). Since today's imperfect recordings typically have both these problems, being too dry and too bright, what can we do in the playback system to compensate for these weaknesses, and bring the music listening experience back closer  to the live concert hall experience? The answer is best found in the design of our chosen speaker system, the design of our listening room, and in getting these two elements to interact productively. The on-axis frequency response of a speaker system should probably be flat, in order to provide you with maximally accurate first arrival direct information. But, as reported in many articles over the years, it has been found  that people prefer speaker systems with an off axis power response that is not flat, but rather is rolled off in the upper frequencies. This can effectively compensate for at least some of the excess brightness encoded in  today's recordings by the mikes' close and high positions. A speaker having flat omnidirectional radiation at all frequencies, or a flat power response, would probably sound too bright (and lean) on most recordings in most listening rooms (you could compensate for such a speaker by  lining your walls with material to absorb some amount of these higher frequencies). Most speakers do have a non-flat power response at upper treble frequencies, where their tweeter's polar response becomes directional. And most  speakers (other than bass dipoles) are omnidirectional at low frequencies, so they have full power response there. Thus most speakers address this desideratum adequately at the extremes of the spectrum. But most speakers, somewhere in the midrange (depending on their crossover frequencies, and depending on whether they're a two way or three way design), have a dip in their power response, where the lower frequency driver (the  bass/midrange in a two way system) becomes directional and where the next higher frequency driver hasn't yet come into play (this smaller driver will establish omnidirectionality again at the frequency where it does cut in, due  to its smaller diameter). Thus, most speakers have a colored, non-flat power response through the midranges. The Samadhi Ichiban sidesteps this problem by making its midranges omnidirectional, splaying them off the listening  room ceiling. Additionally, a speaker interacting with the listening room should utilize reverberation in the listening room to recreate some of the hall reverb and stage ambience that was lost by the necessarily too dry close miking in the  recording. As many articles in IAR and elsewhere have discussed, you can enhance the recreation of the stereo image, especially toward more natural stage depth and hall ambience, by appropriate placement of speakers relative to  room boundaries, and by appropriate treatment of room surfaces. For example, speakers should be placed far enough from the side and front walls so that their first acoustic reflection path to your listening seat exceeds the length of the direct path by at least 10 feet (preferably at least  15 feet), to place it beyond the Haas integration threshold (below this threshold the reflection would muddy the direct sound; above this threshold it can enhance the reverberant ambient sound). The front end of the room should  be treated to be substantially live (not dead), so that a rich reverberant field is generated that can commingle with the direct, too dry signal from the speakers. This added reverberation, created in the listening room, can  simulate the higher proportion of reverberant to direct sound that would be heard from a typical actual seat at that recording venue, rather than from the artificially dry mike location as encoded in the recording, and can  thereby bring the home reproduction of typical recordings closer to the live music experience. Now, the sonic character of live music's reverberant energy at a concert hall is quite dull (even if the direct sound of the music is bright). That's due to many factors: the presence of much absorbent material, from plush  curtains to audience members; the long trips reverberant energy must make between reflections in a large hall (air attenuates high frequencies over long distances); the incoherence of the many random, long path reflections in a  large hall; etc. Thus, if it's appropriate to create some reverberant energy in the listening room to simulate the concert hall listening experience, that reverberant energy should be created to sound quite dull; it should have  full response through the warmth and midranges, but should be rolled off in the trebles. Most speakers, being omnidirectional in the bass and warmth regions, do succeed at feeding your listening room's reverberant field at these lower frequencies. And most speakers with forward facing tweeters, being directional in  the treble frequencies, also do succeed at rolling off the trebles of the reverberant field they create in your listening room. But most speakers leave a hole in the midranges, where their larger driver has become directional, and thus cannot feed the right amount of midranges to the reverberant field created by your speaker at the stage end of your  room. Thus, listening room reverberation from most speakers might have a dark, disjointed, or weak character, not sounding at all like true stage and hall ambience. That's especially true because our hearing mechanism is so  sensitive to the midranges, and we locate and size up objects and spaces most keenly in the midranges, so if the midrange sounds wrong, everything sounds wrong. The Olsher design approach, as in the Samadhi Ichiban, fills in this midrange hole that exists in the reverberant field of most other speakers, thereby making the recreated stage and hall ambience sound more vibrant, more  present, more real. Thus, the Olsher design approach can provide a more accurate re-creation of the natural hall reverberation missing from most recordings, and thus do a better job of re-creating your concert hall listening  experience. Recall that the Ichiban, like the Hegeman, Bose, and Shahinian speakers, puts out a higher ratio of room reflected to direct energy than standard speaker designs. Among these high reflection designs, the Samadhi speakers are  the only ones that put out a spectral balance of reflected energy that closely mimics the character of reflected reverberant energy in the concert hall. Standard forward firing speaker designs put out less room reverberant energy, so they do not re-create as rich a portrayal of concert hall ambience (though some with superb stereo imaging do manage to extract every last bit of  hall ambience that is encoded in a recording, thereby lessening the need to re-create or manufacture any reverberant ambience in the listening room). Some standard speakers have added a rearward or upward firing tweeter to  augment stage and hall ambience, but they still leave an ambience hole in that all important midrange. The Olsher design approach might have one problem area, with its upward firing woofer/midrange driver. At the upper end of this driver's range, where it gets directional, there might be a midrange depression in the frequency  response of the direct first arrival sound from this speaker. In other words, by pointing the woofer/midrange driver upward, the listening room reverberant field gains flatness through the driver's directional midrange region,  but the direct first arrival sound might suffer a hole. One possible tactic around this problem might be to let this driver's on-axis response rise (instead of being flat) as it becomes directional, thus at least maintaining a  flat power response through the midranges. Or, in a more expensive three-way version of the two way Ichiban, a forward facing midrange driver could gradually transition in as the upward facing woofer/midrange becomes  directional. Samadhi's literature mentions that the response of the Ichiban's drivers has been contoured, so we presume that this technical factor has been addressed. The new Ichiban is scheduled to begin production in April. It will be a very strong candidate in its price range, alongside the forthcoming Vandersteen 2ce Signature. We should also mention another Samadhi speaker, the $1395  Magic Cube, at this juncture. Even though it has no low bass, and therefore technically should be rated in a lower class, the Magic Cube shares much of the Ichiban's technology, and has most of the Ichiban's virtues, except  over a narrower frequency range. What makes the Magic Cube truly magic is that its size is so diminutive, a mere 7.5" cube, yet its sound is so big. It projects a big soundfield, thanks to its upward firing woofer/midrange. And its tonal balance is also  that of a big speaker, with rich warmth and upper bass, down to the point where the system dies (nominally 65 Hz). It's amazing to hear such a small speaker have such a full measure of warmth and upper bass (even out in a free  field, away from room boundaries), and therefore be so musically satisfying, as satisfying as much larger speakers. It is also tonally neutral and quite free of colorations over its intended spectral range. Most other speaker  systems this small sound much too lean, thin, and bright. Because it is so small, the Magic Cube is an excellent choice wherever space or decor dictates a small speaker, but where you still want big sound and satisfying musicality. Note though that, for optimum realization of its big  sound, the Magic Cube should be placed relatively far away from room boundaries (including the ceiling), so it is not a good choice for space/decor challenged applications requiting that a small speaker be tucked into a corner,  against a wall, or near the ceiling. Samadhi was also playing a third model at their exhibit, an earlier model called The Experience, housed in a futuristic midi-tower. We found the tonal balance too lean, and the system too colored with driver breakup (perhaps  some enclosure coloration too). © 1998 J. Peter Moncrieff IAR subscription rates are $28 for 12 Hotline issue numbers (1 volume/year), $53 for two volume/years, $75 for three, $100 for four, $125 for five, $150 for six, $175 for seven volumes. For first class USA mailing (much prompter  delivery), Canadian subscriptions, and overseas surface add $10 per volume (12 issues). For overseas airmail add $20 per volume. Back issues contain vital information that's constantly referred to in our present and future  product reviews, and is necessary for understanding these reviews. IAR Hotline is now in volume/year #7 of publication. We suggest you take advantage of the multiple year bargain rate and order a 7 volume subscription beginning  with Hotline #1. Please send subscription requests to: IAR, 1813 Chalcedony, San Diego CA 92109.