|
Time & Phase Coherence
It is with sadness that Thiel Audio closed it’s doors on January 31, 2018. The man behind Thiel Audio was Jim Thiel, one of the rare talents in loudspeaker designs.
His philosophy was that loudspeakers should be time and phase coherent. In very simple terms, what that means is that all the frequencies should reach the listener at the same time. This concept is as close as you can get to a perfect loudspeaker.
The Orion builds on the legacy left behind by Jim Thiel. For the first time ever, a time and phase coherent loudspeaker is available for the diy community.
Orion
Time & Phase Coherent
In designing the Orion, I selected the Audax HM130C0 and the Fountek NeoCD3.5H Ribbon tweeter. The HM130C0 is reputed for it’s fast transient response, which is crucial for time coherence. The NeoCD3.5H on the other hand, has super fast transients. That’s one of the properties of ribbons. The other reason for choosing the NeoCD3.5H is the horn. Due to the depth of the horn, there’s no necessity to tilt the front baffle to align the acoustic centers of the two drivers.
Thiel CS2 2 Step Response
 Fig 1 – Thiel CS2 2 Step Response (courtesy of Stereophile)
In time coherence, the measurement of interest is the Step Response. What we want to achieve is a right angle, triangular response like in the Thiel CS2 2 shown in Fig 1. A step response of this nature indicates all the frequencies are reaching the microphone at the same time.
Orion Step Response
Fig 2 – Orion Step Response. Mic at 1 meter, 1 inch above NeoCD3.5H axis. (courtesy of REW)
The Step Response of the Orion is shown in Fig 2. There is a small dip at the beginning, indicating that the NeoCD3.5H reaches the microphone a split second earlier. It can be eliminated by recessing the NeoCD3.5H back slightly.
The main transient itself is clean, one straight vertical line. What is noteworthy after this leading edge is the tip. In the Orion, there’s no overshoot and ringing. The Thiel CS2 2, on the other hand, suffers some overshoot and rings slightly after that (Fig 1). Because of these, some listeners can find the treble irritating at times.
Now that we have established what Time Coherence is all about, we’ll look into the Phase Coherence aspect. This Phase Coherence is equally, if not more important, than Time Coherence itself.
In Phase Coherence, the important plot to look at is the Excess Phase. Ideally, it should be a perfectly horizontal line.
Thiel CS2 2 Excess Phase
 Fig 3 – Thiel CS2 2 Excess Phase (courtesy of Stereophile)
Fig 3 is the Excess Phase plot of the Thiel CS2 2. Note that there is a slight phase shift. The frequencies above the crossover (about 3kHz) is in negative phase whereas those below are in positive phase.
Orion Excess Phase
Fig 4 – Orion Excess Phase. Mic at 1 meter, 1 inch above NeoCD3.5H axis. (courtesy of REW)
Fig 4 is the Excess Phase plot of the Orion. See how horizontal the plot is. Basically, what this means is that there are virtually no phase shifts from 40Hz to 7kHz. Beyond 7kHz, the high frequencies undergo phase shift progressively. I have not made any attempts to correct this as the Orion is still in the early stages of development.
Thiel CS2 2 Frequency Response
 Fig 5 – Thiel CS2 2 Frequency Response (courtesy of Stereophile)
The frequency response of the Thiel CS2 2 is shown in Fig 5. It is not as flat as one would expect. Most glaring is the -5dB downward slope from 100Hz to 1kHz. This will certainly color vocals. Male voices will sound bassy whereas female ones will lose some presence. The soundstage will be affected too. Female vocalists will recede slightly into the stage instead of standing at the front.
Orion Frequency Response
Fig 6 – Mic at 1 meter, NeoCD3.5H axis. Impulse Window =5 ms. No smoothing. Nearfield below 500Hz.
Fig 6 is the frequency response of the Orion with the same amplitude scale of 5dB as in the Thiel in Fig 5. Note that the Orion is much flatter and the treble extends to 20kHz. The Thiel CS2 2, on the other hand, starts to roll off at 10kHz.
No smoothing is applied in the Orion FR measurement so nothing is glossed over. No active or passive equalization is used to achieve this flatness.
Orion Cumulative Spectral Decay
Fig 7 – Mic at 1 meter, 1 inch above NeoCD3.5H axis. (courtesy of REW)
Fig 7 is the magnified view (1.2 ms time range) of the Orion waterfall plot. Frequencies from 3kHz~20kHz decay quite evenly and quickly. The Fountek NeoCD3.5H is quite outstanding as there’s virtually no artifacts in the high frequencies. Further down the scale, there’s some hotness at 2kHz which is the 2kHz bump seen in Fig 5.
Orion Toneburst Energy Storage
 Fig 8 – Mic at 1 meter, NeoCD3.5H axis. Impulse Window =5 ms.
The Orion’s Toneburst Energy Storage plot of Fig 8 shows clearly that the bumps from 1kHz~2kHz seen in Fig 5 are caused by excess energy (the light blue slices). From 2kHz to 20kHz, there’s no excess energy.
Orion Spectrogram
 Fig 9 – Mic at 1 meter, NeoCD3.5H axis. Impulse Window =5 ms.
The Orion Spectrogram in Fig 9 verifies the excellent performance of the Fountek NeoCD3.5H ribbon tweeter. From 3kHz onwards, it’s already down by -30dB (violet patch) in less than 1 msec. No ringing, no smearing.
the sound of the Orion
This is the first time I’ve ever designed a time and phase coherent speaker and what a revelation. It’s possibly the most natural sounding speakers I’ve ever heard. Perhaps it’s because my brain doesn’t have to continuously compensate for the phase shifts that’s inherent in higher order crossovers.
The Orion would suit listeners that favor vocals and instruments. Willie Nelson, Dean Martin, the Platters, Kenny G, Ben Webster all sound amazing.
In the professional field, it’s ideal as nearfield recording monitors, particularly for recording engineers who are sensitive to phase shifts. You can work long hours with them and you won’t feel mentally drained.
For full range performance, I will have to integrate in a 10″ woofer for the bass. I will keep that as a future project because I want to preserve the time and phase coherence in a 3-way. Meanwhile, I’m enjoying listening to the Orion. |
