*Correction*: 56:18 Equalizer APO uses _IIR_ filters for its graphic and parametric EQs which I believe is inherently minimum-phase in this application, its separately supporting "convolution" filters which are synonymous with FIR. As far as I know, IIR more readily allows manipulations in the frequency domain while FIR takes in a .WAV file of the time domain impulse response and would thus require Fourier transforms to translate frequency domain changes into the impulse response that is being convolved with the audio signal. Likewise, FIR will only cause pre-ringing when you perform _non-minimum-phase_ manipulations (e.g. linear phase EQ) to the impulse response. Else, FIR can perfectly emulate minimum-phase IIR filters, and for a sufficient window length can capture the reflection properties of speaker responses in a room (showing up as delayed spikes within the impulse response, those spikes encoding the spectral balance and timing of those reflections), but necessarily incurring more latency and processing overhead proportional to the FIR filter's window length (.WAV file impulse response window). Those reflection spikes in the frequency domain show up in the sharp comb filtering among other jagged artefacts. The close relationship of the time and frequency domains is quite fascinating. Also, the Meze cables were later found to indeed be much closer than frequency response than as seen in this video, the variations having been due to movements in the headphone pads or in-ear mics. The Meze Elite "V3.1 PEQ" being so flat for my blocked canal in-ear measurements was due to the original Harman EQ based on the headphones.com measurements being similarly flat overall, but being rather wavy due to incorrect adjustments made by ear using sine sweeps and because EQ based on a test rig is not guaranteed to match your own ears. This flatness also lined up with the HiFiMan's expected deviations from neutral. As for the in-ear speaker responses, I later conducted a "threshold of hearing EQ" and determined that due to my in-ear measurements excluding the ear canal, there was actually a major mismatch between the actual at-eardrum response from measuring the speaker response with blocked canal in-ear mics and EQing the headphone measurement to show the same, whereby said threshold of hearing EQ ends up flattening things down closer to the V3.1 PEQ and HiFiMan tonalities but with the "HiFiMan dip" rather occurring around 1 kHz. See www.head-fi.org/threads/recording-impulse-responses-for-speaker-virtualization.890719/page-121#post-18027627 (post #1,812) for more details.
It was the DAC of the MOTU M2 outputting to the balanced line out and into the FiiO K9 Pro ESS's line in bypassing its internal DAC to the THX amp. Listening impressions used the FiiO K9 Pro ESS's internal DAC through its THX amp.
![I.N "HALLUCINATION" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/n5B5q1Hwt_U/mqdefault.jpg)
*Correction*: 56:18 Equalizer APO uses _IIR_ filters for its graphic and parametric EQs which I believe is inherently minimum-phase in this application, its separately supporting "convolution" filters which are synonymous with FIR. As far as I know, IIR more readily allows manipulations in the frequency domain while FIR takes in a .WAV file of the time domain impulse response and would thus require Fourier transforms to translate frequency domain changes into the impulse response that is being convolved with the audio signal. Likewise, FIR will only cause pre-ringing when you perform _non-minimum-phase_ manipulations (e.g. linear phase EQ) to the impulse response. Else, FIR can perfectly emulate minimum-phase IIR filters, and for a sufficient window length can capture the reflection properties of speaker responses in a room (showing up as delayed spikes within the impulse response, those spikes encoding the spectral balance and timing of those reflections), but necessarily incurring more latency and processing overhead proportional to the FIR filter's window length (.WAV file impulse response window). Those reflection spikes in the frequency domain show up in the sharp comb filtering among other jagged artefacts. The close relationship of the time and frequency domains is quite fascinating.
Also, the Meze cables were later found to indeed be much closer than frequency response than as seen in this video, the variations having been due to movements in the headphone pads or in-ear mics.
The Meze Elite "V3.1 PEQ" being so flat for my blocked canal in-ear measurements was due to the original Harman EQ based on the headphones.com measurements being similarly flat overall, but being rather wavy due to incorrect adjustments made by ear using sine sweeps and because EQ based on a test rig is not guaranteed to match your own ears. This flatness also lined up with the HiFiMan's expected deviations from neutral.
As for the in-ear speaker responses, I later conducted a "threshold of hearing EQ" and determined that due to my in-ear measurements excluding the ear canal, there was actually a major mismatch between the actual at-eardrum response from measuring the speaker response with blocked canal in-ear mics and EQing the headphone measurement to show the same, whereby said threshold of hearing EQ ends up flattening things down closer to the V3.1 PEQ and HiFiMan tonalities but with the "HiFiMan dip" rather occurring around 1 kHz. See www.head-fi.org/threads/recording-impulse-responses-for-speaker-virtualization.890719/page-121#post-18027627 (post #1,812) for more details.
Hi, what DACs and AMPs were you using ?
It was the DAC of the MOTU M2 outputting to the balanced line out and into the FiiO K9 Pro ESS's line in bypassing its internal DAC to the THX amp. Listening impressions used the FiiO K9 Pro ESS's internal DAC through its THX amp.