In this blog post I test the 18Sound ND1480BE Beryllium compression driver. This driver retails for €1,350.00 each ($1,485 USD).
Features
- 3 inch premium Beryllium dome
- polymer surround diaphragm
- Extended HF with minimal breakup
- 112 dB 1W / 1m average sensitivity
- 1.4 inch throat exit
- 3 inch edgewound aluminum voice coil
- HF copper sleeve for reduced distortion and increased output
Published Data
Below is the published data from the manufacturer. The impedance sweep shows the FS centered around 460Hz with the first breakup showing up at 17kHz. 
The published response shows the typical CD Horn falling response. The horn used in the published data was the XR1464C.
My Own Measurements
For this test I used the ES-600 Biradial which is shown below. The horn uses a standard 1.40" throat diameter with a 600Hz cutoff Frequency (Fc).
Impedance
Below is the measured impedance and phase response for the ND1480BE. The mechanical resonance (FS) is centered at 460Hz which closely matches published. The first mechanical breakup occurs at 16.5kHz
The first mechanical breakup occurs at 16.5kHz as you can see below.(zoomed)
Below is the measured frequency response with an 8.2uF capacitor (green), 3.9uF capacitor (blue), and 1.50uF capacitor (purple). 
Time Domain
With the mic positioned at 30cm, I measured the burst decay using a 5.66ms gate. I also used the 3.9uF capacitor to flatten the response somewhat for this measurement.
Below is the CSD plot. We can see that
Harmonic Distortion
Below is the harmonic distortion result for the 85dB test signal level at 1m. The mic was placed 15cm from the horn mouth. H2 remains below 0.10% for most of the spectrum with H3 and H4 at 0.015% and 0.00613% respectively. 
Raising the test signal to 95dB we see a modest increase in H2 with H3 and H4 remaining very low at 0.037% and 0.00685% for the 600Hz region. 
If looking at the same results but with the vertical scale changed to dB instead of percent, we get the following for the 85dB test signal level.
For the 95dB test signal level we get the following.
For comparison's sake I've shown the Eminence N314X below.
Intermodulation Distortion
I then measured intermodulation distortion using a 12 band per octave test signal ranging from 600Hz to 20kHz set to 85dB SPL at 1m.
Raising the test SPL to 95dB is shown below. 
I decided to limit the test signal to 1kHz and above. This should reduce the amount of modulation into the upper frequencies generated by the lower tones.
Again, at the 95dB test signal level.
Subjective Listening Impression Ranking
With the speakers setup as a stereo pair, I evaluated using a 2-way setup. Bass frequencies were handled by the Fostex 8" (FW208HS) in a 40L cabinet tuned to 35Hz. The high pass filter was a single capacitor (3.9uF) followed by a fixed resistor L-Pad comprised of a 6.8ohm (R1) and 1.5ohm (R2). This provided -16dB of attenuation to match the 90dB sensitivity of the Fostex 8".
Soundstage Depth --- 9/10
Soundstage Width --- 8/10
Smoothness --- 10/10
Coherence between midrange and treble --- 10/10
Vocal Clarity Male --- N/A mostly evaluated 600Hz and above
Vocal Clarity Female --- 10/10
Accurate Musical Instrument timbre --- 10/10
Sense of Dynamic Range: --- 10/10 Provided the same dynamics as titanium but without the metallic sound signature.
General Thoughts on Sound Character
The ND1480BE surprised me for its ability to reproduce upper frequencies with precision and accuracy. Cymbals and electric guitar was faithfully rendered in terms of texture and nuance, but also with proper dynamics. The driver does not sound soft, but exerts itself with confidence and grace. It also did not come across as harsh or metallic in any way. The midrange and treble offers very good clarity. As I play track to track, each recording takes on a unique character that is colorful and engaging.
Context
My subjective assessment is in the context of medium size format horns such as the ES-600 Biradial. If we are having a general discussion about ultimate sound quality, then it needs to be pointed that a larger horn will provide a larger sense of scale and clarity, and in turn a more engaging listening experience. If we talking cost-no-object ultimate sound quality then the ES-450 or ES-290 Biradial with the TAD TD-4003 will provide those attributes mentioned.
Comparison to SB Acoustics 65CDN-T
Directly comparing the two drivers reveal that they are closer than one might expect. They both offer the same perceived frequency response balance, however the 18Sound offers slightly more soundstage depth due to the slight increase in clarity. Both drivers share the same smooth character. The differences between the two drivers are not as significant as other potential variables in the system such as source sound quality. In other words, these differences could easily be masked by limitations in the upstream components, just to provide some context.
18Sound Beryllium or 65CDN-T with T500A Mklll?
Rather than spend the money on the 18Sound it occurred to me that for a little more money you could buy both the 65CDN-T and the flagship Fostex T500A Mk lll super tweeter. Would this combination be preferable in terms of sound quality? To test this I setup both and directly compared.
The 65CDN-T + T500A Mklll combination provided a little wider soundstage width but the real difference was in the characteristic alnico sound of the Fostex. It has a slightly richer sound character in the treble region compared to the 18Sound, where the 18Sound comes across slightly grey sounding. From a measurement perspective I would suspect this is attributed to the higher level of H2 harmonics with the Fostex. To show this, I've shown the T500A Mklll harmonic distortion sweep for the 85dB test signal level. As you can see H2 is over 1% for a test signal which represents normal listening levels.
So this begs the question as to weather this H2 attribute could be added with a good tube preamp when using the 18Sound.
To test this theory I added the Don Sachs Tube Preamp to the system which is shown below.
I had conducted a full set of measurements on the Don Sachs Preamp a few months earlier using my electronics test bench.
Below is the harmonic distortion at 4.00V line level output for the Don Sachs Preamp. As we can see there is a ladder effect where H2 is predominate followed (blue line) followed by H3 and H4 at the lower rungs.
Setting this up and listening confirmed what I suspected based on the measurements. The Don Sachs preamp brought a romantic sound character to the entire frequency spectrum (not just the treble) for the 18Sound compression driver. This became especially apparent on Mary's Lullaby, Cantate Domino is 24bit 88kHz HD track...
This is more of an academic interest which is moving beyond this compression driver review. But essentially the 18Sound offers a clean slate to which you can "flavour" the sound character to your liking with the addition of a tube preamp or other method (DSP?).
At this point I would place a tie between the two configurations. The addition of the tube preamp leveled the playing field against the 65CDN-T + T500A Mk lll.
The question then becomes...what happens when I use the tube preamp with the 65CDN-T + T500A? This configuration was even better than the 18Sound + Preamp. The Don Sachs preamp has a way of widening the soundstage even further versus no preamp with the T500A mklll. There becomes more space around the vocalist. The romantic aspect is increased as well as a result of the above attributes.
This could also serve as a caution when evaluating compression drivers. If the driver sounds cold and clinical, it could be the lack of H2 distortion and not some sin of commission from the compression driver itself.
Conclusion
My listening impressions on the ND1480BE left me impressed. The driver provides impeccable measurements as well, especially in the 6kHz-12kHz region.