I finally purchased a pair of M22ti's. I had a chance to compare them against some speakers I own.
All 4 speakers were designed using data collected from years of NRC research into sound reproduction, yet all 4 sound quite different from each other.
The Contenders (and their Stories)
1) Paradigm Phantom (1993 - $330 Cdn)
My first audiophile speaker, this speaker earned an NRC 7.3 fidelity rating in Sound and Vision (Canada). I auditioned many speakers before buying the Phantoms. Compared to various Boston Acoustics, Infinity, Cerwin Vega, and Acoustic Research speakers, this was an amazingly neutral speaker.
Afterwards, looking for a reference speaker, I bought a pair of $2500 Polk Audio LS90 tower speakers, before realizing that the Phantoms were more neutral than the Polks. The Polks were returned. Not wanting to be burned again, I was fortunate to find a dealer who let me bring my Phantoms to the store for comparison. Unfortunately for him, my Phantoms laid waste to about 6-7 pairs of $1000-$3000 speakers I had considered, and only the $3000 NHT 3.3 was noticeably better (mainly in the width of the soundstage).
The sales literature of that model listed specs of 60-20kHz +/-2dB on axis.
2) Nuance 330 (1995 - $3000 Cdn)
The speaker to beat in my house.
Nuance Audio is a company that makes PT Barnum seem like Mother Theresa. Nuance is found only in seedy electronic chains that cater to the Chevy vs. Ford and show-us-your-tits crowd, a group that knows nothing of the phrase “deferred gratification“ and are financially known as “sub-prime”. The “Nuance Nuke” sales pitch using Q Sound recordings is so effective at suckering them in, Edmonton and Calgary readers will find dozens of them for sale in local bargain finder newspapers or pawn shops. I briefly worked for a store that sold these speakers, and even though the 330’s sold for $3000, store cost was approximately $600 a pair. That’s 500% markup, which apparently is the norm in the retail speaker industry.
Even then, I knew the flagship 330’s were something special. Noticeably better than the rest of their line, I had the fortune to compare the 330’s against the $6000 Thiel CS 3.6 within an hour of each other, and my friend and I came to the conclusion that the 330’s had the same midrange but much deeper bass than the Thiel. The fact that there were being discontinued meant that my friend and I each picked up a pair for $550.
After buying them, I continued to explore a few other brands. The $6000 KEF Model Four seemed to be it’s equal, so I decided that the 330’s would be a Buy and Hold speaker. My Phantoms were relegated to computer speaker duties. Every once in a while I’d audition speakers such as the $10000 Meridian DSP5000, $12000 KEF Model 205, and $5000 Revel Performa F50, but I usually came to the conclusion that they were no better than mine.
Yet I had my doubts as to its accuracy. Listening to certain tracks, sometimes I feel the lower treble is harsh, and I wanted to find out if a better speaker was available for a reasonable price. I also wanted to move away from a full size tower speaker to something that could be later wallmounted.
3) Paradigm Titan (1999 - $200 Cdn)
I won a $200 gift certificate and some CD’s (blues/jazz, yuck!) at a local audio salon, and wanted a pair of computer speakers that were less unwieldly than its big brother Phantoms.
The Titans are legendary. In a Tom Nousaine article in Sound & Image magazine 1994, the Titans soundly won a 20 way bake-off of name brand bookshelf speakers. After that, no other audio magazine wishing to stay solvent ever did fair comparison tests again, and Sound & Image folded around 1996. A home theatre system featuring the Titans were given an NRC fidelity rating of 7.2 in Sound & Vision (Canada) magazine.
The Titan sales literature specs performance as 60-20kHz +/-2dB on axis, the same as the older discontinued Phantoms.
4) Axiom M22ti (2005 - $530 Cdn)
The M22ti has big shoes to fill. Upholding the honor of it’s big brothers M60ti and M80ti, it is considered a timbral match to it’s larger brothers. Combined with the reputation of this speaker line to equaling the Paradigm Reference, Energy Veritas, and PSB Stratus series at a fraction of the price, this speaker had a good chance of replacing my 330’s in my family room.
I realized early that volume matching would be difficult. The Titans were the most efficient, followed by the evenly matched Phantoms and the M22ti‘s, and the 330’s the least efficient at 1-2 dB down. The 330’s would be at a disadvantage during testing. The Axioms had little bass, so to make things fair I augmented it with my sub during testing against the full range 330‘s. I would listen A/B with my integrated amp, but later brought in my Kenwood receiver to take advantage of its nebulous A/B speaker switching that would approximate a reasonable single blind test.
I would be switching A/B on various familiar song tracks, and keep track of which was better. I would start off comparing them in freespace at first, and then next to a back wall. I would give myself as much or as little time as needed.
Round 1: Paradigm Phantom vs. Paradigm Titan
It became obvious that both were timbrally matched. By themselves, both sounded great. Depending on speaker placement, I either preferred the Phantoms in free air or it was a tie. The Phantoms were a bit smoother in the midrange, slightly deeper bass and less shouty/edgy compared to the Titans. The Titans were more forward, and seemed to have a more aggressive tweeter. The Titans definitely need a subwoofer. Otherwise, both were similar enough to one another to be almost a tie. Depending on program material, the Phantoms would pull away, particularly on Sarah McLachlan tracks.
Round 2: Paradigm Phantom vs. Nuance 330
On female vocals, the Phantoms had a shouty, edgy, congested midrange, and slightly less bass compared to the 330‘s. Male vocals were better, but the same colorations were there. The 330’s simply disappeared, as if listening live. On every music selection, I preferred the 330’s. Amazing what happens when you change reference groups.
Round 3: Nuance 330 vs. Axiom Audio M22ti + Mirage PS12/90 sub.
Listening to the M22ti, I was always aware that I was listening to loudspeakers. The 330’s midrange seemed a bit more laid back in midrange compared to the other 3 pair of speakers, but it was always crystal clear and smooth, with a wider soundstage. The M22ti’s comparatively sounded tunnelly or cuppy, and compressed at moderately higher volumes. The harshness I thought the 330’s would have in the lower treble simply wasn’t there, or if it was, the M22ti didn’t help reveal it. Playing around with speaker locations, the M22ti’s proximity to the 330’s affected its sound negatively, but not vice versa. On every CD track, I preferred the 330’s to the M22ti.
Round 4: M22ti’s vs. Paradigm Phantom
Both speakers sounded fine. While not sounding alike, the phrase “similarly good” would be the best description. Depending on program material, my preference would flip back and forth. The Phantoms had more bass, but that advantage was negated by a sub.
Round 5: M22ti’s vs. Paradigm Titan
The M22ti’s were noticeably smoother in the midrange and treble. The difference was more pronounced than when comparing the Titans against the Phantoms. In every instance I preferred the M22ti.
The M22ti’s were very respectable. I felt bad about returning the speakers, since only the courier companies benefited from this transaction. However, if the midrange is king, then they lost to a significantly more neutral and natural sounding speaker. Rather than start auditioning speakers priced like new cars, I think I’ll just enjoy what I have for the next 5-10 years.
Concerning the test methodology, I concluded that:
1) You get used to a speaker’s sound very quickly, warts and all. Only when comparing against another better speaker will flaws become noticeable.
2) A/B switching is a more reliable way of detecting colorations than trying to listen at leisure. It only takes at most a minute or two to decide which is the better speaker. Acoustic memory is unreliable.
3) Speaker placement matters. Some speakers are more sensitive to placement than others.
4) The more accurate speaker becomes the baseline. The better your reference speaker, the easier it is to pick out colorations and deficiencies of the less accurate model.
5) The louder speaker does indeed sound better, if you weren’t aware of it during testing. That’s probably why the Titans did as well as it did, its 1-3dB advantage over the other models making it sound better than it really was.
Nuance 330 (1st)
Axiom M22ti (2nd - tie)
Paradigm Phantom (2nd - tie)
Paradigm Titan (4th)
[Test Equipment: Denon PMA-915R (80W/ch @8Ohm, 130W/ch @4Ohm), Kenwood 1080VR, Pioneer DVL-909 DVD/LD player, Mirage PS12/90 sub, 16 gauge speaker wire, cheapo interconnects]
[Music: Female vocals: Le Couple, Noriko Sakai, Shirley Kwan, Sarah McLachlan, Holly Cole, Rankin Family. Male vocals: Junk, Ange Tong, Michael Buble, Rankin Family.]