Eminence N314X on Horn No.1876 (ES-600 Biradial)

Eminence N314X on Horn No.1876 (ES-600 Biradial)


In this blog post I test the Eminence N314x 1.40" compression driver featuring a 3" Textreme diaphragm. I will provide full test data, with and without the rear cover, as well as my subjective listening impressions for a 15" 2-way application using a 600Hz crossover point. 

Why test the N314x?

I received the driver as a sample from Eminence themselves after I had requested a sample. For the past few months I've been getting emails from customer's requesting that I review the driver. The last time I tested a Textreme driver was the Lamar M2-16 Field Coil unit from Germany. I enjoyed it's particular tonal character which I found to be rich and warm. Was this attributable to the Textreme diaphragm material? If the Eminence has a similar sound character as the Lamar, then perhaps this may be the case. 

For interest's sake, Audioxpress has also tested the driver which can be found here. They tested the compression driver on the Eminence H14EA horn. 

Also, Nick Toid from Nick Toid's Audio did an interview with Matt Marcum, Senior Design Engineer from Eminence which can be found here

Objective Test Data

I begin by conducting an impedance sweep using my DATS V3 tester. We see that the FS is at 464Hz. 

I then measured the on-axis frequency response using a 60cm mic distance. Things are generally flat until we get to the two peaks at 7kHz and 8.6kHz which deviate +/-3dB. We then see the response shelf down -5dB at 10kHz. But the general behavior in the 10-20kHz region is free from what we would normally see in a metal driver...which is severe breakup which would normally introduce massive peaks in the response. It's common to see +20dB peaks in this region. Instead what we see is a falling response out to 20kHz with deviations of around +/-3dB. It's not until we get to the 20-30kHz region do we see some severe breakup. 

Time Domain 

Burst decay shows resonances extending beyond my target 8 periods starting in the 7kHz region. 

The CSD plot shows a reverse null at exactly 8kHz that extends a little past 1ms. This high Q resonance appears to the the only resonance that extends past 1ms within the audible band. Otherwise decay settles in under 1ms. 

Comparing against published we can see the two peaks at 7kHz and 8.6kHz are the longest decay artifacts coming to rest just before 1ms. My data shows this at around 1.4ms which is still good. This translates into about 12 periods in the burst decay using a 35dB vertical scale. For me, I'm not sure if this would be audible. It certainly did not detect any coloration in this region during my listening test. 

I changed the settings in ARTA to have the same time scale (see 6.12ms above) to see an apples to apples comparison of my data versus published. By changing the FFT Block Shift in Samples from 2 to 5 we can see the time scale increase to 6.69ms (see below). We can see the graph closely matches Eminences published data. 

For comparison, I've kept the CSD plot settings the same as above and looked at the result for the SEAS Prestige 27TBFCG (H1212)  Mounted on Horn No.2007 (see below). Generally speaking, when comparing data you must look at the graph scaling for a fair comparison. 


At 85dB test signal we see H2 remain low at -68dB (1kHz). H3 and H4 are below the noise floor of my measurement system which is -90dB. 

Increasing to 95dB we see H2 rise in a linear fashion to -58dB.  We see a slight rise in H2 in the 8kHz region, rising +4dB compared to other areas. 


I tested the driver at 85dB which shows acceptably low IMD. The midrange is at -72dB rising to -65dB at 10kHz.


Increasing the test SPL to 95dB we see distortion rise in a linear fashion. At 10kHz we see IMD at -55dB. My in-house target is -60dB which was achieved here as an example, but I am splitting hairs at this point. 

Subjective Listening Impressions 

I set up one speaker in a 15" 2-way configuration for my evaluation. Below is a summary of the test steup... 

Laptop (Plex Software) >Topping D10S Digital Out > Hypex FA123 Digital Input > Speakers 

Below is the EQ settings I settled on in the Hypex Filter Software. 

Below is the resulting frequency response for the above EQ settings. 

Soundstage Depth --- 9/10  Excellent low level detail retrieval 

Soundstage Width --- N/A (listening in mono) 

Smoothness --- 9/10 Similar to most soft dome tweeters 

Coherence between midrange and treble --- 8/10 

Vocal Clarity Male --- N/A mostly evaluated 600Hz and above 

Vocal Clarity Female --- 8.5/10 

Accurate Musical Instrument timbre --- 9/10 

Sense of Dynamic Range: --- 7.5/10 The slightly softer sound character came at a compromise against dynamics. 

General Thoughts on Sound Character

When I first started listening to the driver my immediate thought was that this compression driver sounds similar to a soft dome tweeter in it's overall sound character, in that is actually sounds a little 'soft'. The driver responded well to EQ as I introduced some subtle -3dB notches at various areas through the midrange. Bringing up the treble helped flesh out and identify instruments more clearly in that region (maraca, tambourine etc) to the point where I didn't feel I needed the Fostex super tweeter, although the addition of the T96A still helped bring out a little more shimmer. The soft character provided some relief to harsh soundtracks. 


I found the driver to sound pleasant and musical with virtually no listening fatigue. Despite the test data showing some anomalies in the upper treble, I found soundstage depth to be excellent across it's bandwidth. The driver was able to fully render acoustic guitar, banjo, and the ukulele correctly. Even high treble instruments such as the maraca and tambourine came through convincingly accurate, despite the slightly soft sound character. The Eminence N314x seems to be a good all-around performer in terms of balancing out all of its parameters to provide a pleasant and enjoyable home listening experience.

Back to blog