• Sonic colorations are particularly noticeable on simple vocal and instrumental music (e.g. guitar) where individual notes are well separated (frequency sweeps were especially revealing in this regard). To what extent the colorations impaired the music varied according to the recording but when the conditions to arise were ideal for them, it could be distracting.
• Discontinuity between bottom and upper registers. How integration is wanting was dissimilar from how the tweeter/mid transition can be discontinuous in a conventional two-way speaker with a poorly designed crossover. What separates the two bands in the Cornu is their textural difference, which separates rather than connects these areas.
• Finally, the excessive sound energy around the boxes seems reinforced by the wall. The listener senses, more or less consciously, that too much is going on there. This feature doesn't really affect the music per se but can blur imaging to make the sound a little phasey and less cleanly articulated than it could be.
  

To say that these problems exist is of course to say that they existed in my room. To avoid or minimize their influence is absolutely essential to find a setup in which these speakers perform at their best. It's a further must because on the wall, the speakers are in an ideal position to excite all the room modes. Once the best spot on the wall has been fixed, the thing to determine is correct listening distance. So let's move to that next.

Daniel's permission) take my review pair to the main DIY event in Finland's Tampere to see how they would compare with small Fostex, Jordan and Tangband projects. It turned out to be impossible to make any meaningful comparisons but the trip proved not entirely useless because the room where the Cornu CS ended up on audition was spacious. Although some upper bass accentuation remained recognizable, the overall sound was more open, coherent and ultimately more believable than I managed to achieve at home. The sound at Tampere reminded me remotely of how the Cornu CS fared at ETF2004 (albeit such past aural comparisons are not very interesting). This is, of course, as it should be since even smaller speakers typically sound better in larger rooms.


Back at home and basically finished with my listening sessions, I still had a nagging suspicion that the sound wasn't as good as it could be and that the reason for it didn't rest exclusively with the speaker. So I decided to conduct some further experiments reminiscent of Mr. 'Wharfedale' Briggs' experiments with classroom speakers. I first placed one speaker in the corner of the short wall at the height of 175cm (measured to driver center) and the other about 1.5 m away on the long wall, i.e. in a 90° angle vis-à-vis each other. Not bad at all! For some of the music I used for critical listening, this obtained the best results in my room thus far, one explanation being that different room modes excited by one speaker were cancelled by the other.

resonances to the sound or if it does, it is well documented. The results of this trip could be summarized as follows:

• The upper bass/lower midrange performance was duplicated by my own earlier results both subjectively and objectively (measurements), meaning that there were certain irregularities in the response over this frequency range.
• When the speakers were moved away from the wall (1.25m), the response retained its characteristics but the midrange became elevated, clearly proving that the speaker was designed for wall placement.
• Subjectively speaking, the most pronounced difference compared to my listening room was some harshness in the upper midrange resembling intermodulation distortion. This was particularly striking on vocal music. This phenomenon didn't go unnoticed at home (for example, van Morrison's voice produced such harshness) particularly when the listening volume was high.

• Curiously, it became much more pronounced in this room than mine and I still wonder whether the high-power transistor amp on hand really was the best match.
• Finally, the sound was lively, energetic and very extrovert, causing no mental efforts to decide what the music was all about.

Listening to the Cornu CS in the near field wasn't enough for best results. I soon found out that it's also very critical to use correct listening volumes. Fixing the playback level for each track and with regard to the capabilities of the room and equipment is a vastly underestimated issue in current HiFi discourse. With the exception of Jean Hiraga of La Revue Du Son, I don't know anyone else who has seriously addressed this issue. With the Cornu CS, correct playback levels are of the utmost importance. Recordings that are critical in this regard (cembalo for example) can be spoiled if played too loud. Because they are so critical to the volume level, some recordings with the Cornus may not be playable at the level they should be in order to maintain timbral qualities. This, however, is compensated for by the fact that sonic expressiveness does not collapse when the volume is turned down. This was a very positive feature of the Cornu CS indeed.

Said dip was clearly visible in all the measurements I made. Judging from the window of this depression, it could be explained by diffractions and internal cabinet radiation. Be that as it may, sometimes the valley was deeper, sometimes shallower. Sometimes it was wider, sometimes narrower. But even at best, it was deep and wide enough to potentially affect the playback. Did I hear it then with music?


At first, I didn't pay much attention to it. Slowly I began to feel "missing presence" on certain music. Philips has an LP on the back cover of which the average distribution of different frequencies during a cut is given in graphical form. A striking feature of the two piano pieces contained on the LP? There is hardly any energy above 2KHz. It was with this sort of music material -- piano in particular -- that the suckout (roughly between 600 and 2000Hz) made the sound a little uninspiring. While casually listening to the Cornu CS otherwise, it wasn't too easy to point the finger to the exact effect of this frequency domain depression.


Not being entirely happy with what I was hearing or able to hear, I decided to do what Dr. Green from TAS would have suggested I do, namely use a reverse method: see what happens if one leveled off the frequency response. So I hooked up a Behringer 1024 digital equalizer between my CD transport and DA converter. This is where such an equalizer should be inserted for minimally deleterious effects. It should be stressed that the aim of this effort was not a perfect equalization of the in-room frequency response. That would be a different task requiring more sophisticated equipment. Rather, by switching between the equalized and straight sound, I tried to trace down what music would be most affected by the aberration. There are very few fundamentals of any instrument in this region so the question became, what were the instruments whose most characteristic harmonics would hit this region? My choice for the test became piano, soprano, violin, cembalo and flute.

Piano fared as expected: there was a slight alteration in timbre and the lower end of the suckout affected the sound as well. Equalization affected soprano voice less than expected and in fact, the Behringer had a greater impact on male baritone voice. Violin has important harmonic overtones in this region and there was indeed a change with the EQ but the potentially most timbre-altering frequencies were well outside the suckout window. It truncated the cembalo's tone a bit mainly at the upper end. The flute remained essentially untouched. In conclusion then, the effect of this frequency irregularity on music was real but not obvious. The essential question is not whether one can hear it but how it affects the music. Even if one can discern it, music can still sound and feel good.


With the Behringer in the system, I couldn't resist to dial in some equalization for the two octaves below 600Hz. I became convinced that using proper equalization techniques (DSP with phase correction etc.) could be a very good idea not necessarily because the speaker requires it but because such corrections could greatly minimize the room interaction problem.


Amplifier control
Many manufacturers stress the importance of marrying their speakers to high-quality amplification. Every time I see that, my first thought -- given that the aim ostensibly is not to explain away certain design faults -- is that the manufacturer means to brag about how revealing his speakers are. I'm not saying that amplification is not a significant issue. On the contrary. But there's also hype involved. Still, my experience of single-driver speakers is that they are able to reveal the tone color built into an amplifier to an extent that is less common with other type of speakers.


Because of their sensitivity -- they sound far more efficient than their figure of 91 dB/W/m indicates -- and fairly easy load impedance, the Cornu CS is suitable for low-powered tube amplification. Consequently, I first mated the Cornus to 10-15 watts (according to certain distortion figures) of my PP amps (a 6V6 vintage and a Shindo Montille EL84), and then to 4 watts compliments of the Sun Audio 2A3 SET (thanks, Tapio). The speakers also got a taste of a 200wpc Densen B-330 power amp (thanks, Juha) and my workhorse, a 200-watt Sony integrated, which I mainly use for measurements.


The Cornu CS was less revealing of midrange timbral nuances than other single-driver speakers I have heard. Not that I couldn't detect the differences in tone -- or traces of them -- I normally associate with well-executed PP and SE amps or between the latter and SS amps, for example. But nothing extraordinary came from such considerations. Neither of the solid-state amps caused extra harshness in the upper midrange. When there was such harshness or congestion, it took place with both types of amps even though I now think that with tube amps, it was easier to avoid.


This doesn't mean that amplification would not be an important issue with Cornu's Compact Spiralhorns. It is but in a more traditional manner. It once was common practice to play with the damping factor of an amplifier in order to regulate the speaker's bass performance. In order to boost bass performance by a few decibels at the resonant frequency, one could insert a series resistor in the voice coil circuit of the speaker and thus reduce the damping factor of the amplifier (a very handy trick with small open back baffles).

But first, speech. Speech reproduction, although calling for somewhat different properties than music, is a good revealer of sound quality. A speaker that fails the speech test rarely satisfies with music. Every now and then, I settled down to listen to familiar radio programmes with familiar voices. The Cornu CS passed the test fairly well. The only major complaint I had was some kind of choking in the lower register of human voices. Such voices didn't quite open up as they should.


With music, the first property I want to mention is a certain compactness of the sound. Putting aside occasional colorations in the bass & lower midrange, the sound had healthy uniformity - an ability to hold musical events together. A good example of this was big band music: whenever the ear was exposed to it, it needn't fear distracting anomalies.


The sound was compact also in a literal sense. Piano in a concerto could take on the tinkling sound of a children's piano and a similar 'narrowing' took place with some violin recordings. I know that friends of tiny wide bandwidth drivers fiercely disagree with me but presenting instruments with sufficient body and timbral maturity including their natural or semi-natural size is one area in which I doubt that a four-inch driver can achieve the same success as, say, 8 to 15-inch drivers are able to. But this is only my speculation.


Am I inconsistent now? I claimed earlier that the Cornus can have a surprisingly big and effortless sound. And so they can. They can easily fill an average-sized room with sound of notable width and height. They convey a feeling that the sound is everywhere in the room (except for on your face). Because of this characteristic, they offer a huge scenery for opera singers, for instance, to play out their roles. In this regard, there is something special about how these speakers load the room. This is not at all incompatible with the fact that occasionally, certain instruments will exhibit a miniaturized sound.


Thirdly, these speakers are fast and agile, with no sense of delay or loitering. No lost sense of time and impeded tempi here. However, the speaker does not seem to be able to follow tinier variations in speed -- those that separate different interpretations from each other -- to the same extent that the best speakers can. The same is true of transient response, the speaker's reaction to dynamic variations in music. The CS can go from quiet passages to fairly loud quite easily but when it comes to smaller variations (e.g. from pp to ppp) their performance is rather approximate.


Fourth, the sound has a quality that could be described by such attributes as lively, unforced, unhindered, effortless. Orchestral music in general comes out easily and in a relaxed manner but baroque orchestras are this speaker's real bravura. Try something like J.S. Bach's Concertos for Cembalo and Orchestra e.g. BWV 1054. You'll be surprised. Also, I'm one of those who claims to love Beethoven's string quartets, even his later ones. What I require from the sound is absolute directness, liveliness and lack of any kind of suppression or inhibition. The Cornu CS did support these quartets and other chamber music in exactly this way.


Fifth and in absolute terms, the Cornu is musically not the most transparent. By this I mean that faint musical events were not as palpable as they could have been. They didn't gently force the listener to pay attention to the smallest details which is important for example with many contemporary music rich in color and effects. I found them to be better with music that doesn't suffer too much from the fact that it is served up in somewhat more enveloping rather than dissective manner.

Sixth and not unrelated, these speakers have what Jules Coleman would call a 'holistic' sound. To put it less sophisticatedly, parts are not for sale. This is not the kind of speaker that, despite of its somewhat dichotomized performance, is chosen for one or two of its characteristics knowing the price one would have to pay in other performance areas. One consequence is that the Cornu becomes insensitive to different source material: what sounds good sounds good, what sounds wanting sounds wanting. The more one listens to them, the more predictable they become. It follows that a test-CD/LP with variety of music offers appropriate material for assessing this speakers' musical competence.


Here's what happened with High Fidelity's Sample Record No. 54 (with the speakers' centers 1 meter from the floor and the listening distance between 1.5 and 2 meters):