How to calibrate a personalized HRTF measurement system

[Mr. Haelscheir]

I am looking for advice on the practicality of calibrating the in-ear microphones and electronics that I intend on using for conducting in-ear headphone and speaker HRTF measurements. I desire reasonable accuracy from 20 Hz to 20 kHz, and though these will of course be specific to my own auricles, I desire advice in developing an appropriate calibration reference for these measurements if anything beyond a channel match is even worthwhile.

The microphone to be used is https://www.digikey.ca/en/products/detail/pui-audio-inc/TOM-1537L-HD-LW100-B-R/12152294 which will be embedded at the end of a 1 cm canal impression/plug as a more comfortable alternative to the Earfish in-ear mics used in https://www.head-fi.org/threads/mez...eadphone-official-thread.959445/post-17743502 (post #5,152). I may have an additional pair through which I will attach probe tubes via David Griesinger's method for speaker HRTF measurement and to hopefully avoid the trouble of having to use hearing threshold EQ to compensate the response as covered in https://www.head-fi.org/threads/rec...-virtualization.890719/page-121#post-18027627 (post #1,812). From what I understand, probe tube mics can be calibrated by having a reference mic situated next to the probe tube's end, though I am still researching this.

The audio interface is a MOTU M2 to which the in-ear mics will be connected with RØDE VXLR+ adapters unless I find a lower-noise implementation (see https://audiosciencereview.com/foru...hones-with-motu-m2-and-rew.49384/post-1783062 (post #6)). The MOTU M2's DAC line out will be used to drive either the FiiO K9 Pro ESS's 4.4 mm balanced input for headphone measurements or the Genelec 8341A with GLM's analog input for speaker HRTF measurements.

The measurement capture software will include Room EQ Wizard, the https://www.earfish.eu/ HRTF capture software, and Impulcifier.

Crux: For both the microphones and the ADC and DAC electronics, there will exist frequency response nonlinearities I would ideally like to correct with calibration files to be linear from 20 Hz to 20 kHz. There are a few local calibration services within Toronto, Ontario, Canada, but as far as I can tell, these only service professional applications with lab-grade equipment as most likely would not offer services for calibrating mere 4 mm electret microphones or a consumer audio interface (or they would only offer services for calibrating the professional equipment that you would use to calibrate those). I suppose I could go ahead and acquire and trust an Earthworks measurement microphone to conduct free field and if possible pressure field calibrations of the electret microphones, but this leaves in the nonlinearities of the electronics. Likewise, I would need to supplement the DIY calibration with a good subwoofer if I wish to ensure that my measurements are accurately capturing the headphones' bass extension.

Hence:

1. Are there recommended services I may not have the right search term for for calibrating these electret microphones and electronics?
2. For in-ear HRTF measurements, blocked canal and at-eardrum probe tube measurements respectively, is a free field or pressure field calibration required? Does this even matter for non-academic and spatialization uses?
3. Is there no way around supplying my own subwoofer to calibrate mic linearity down to 20 Hz?

The simple alternative is to just choose the electret mic for which my Genelec and headphone measurements look the most linear and create calibration files to match the other mics to that, but all measurements would have to be relative rather than absolute.

 

[Speedskater]

You might look at what hearing aid professionals call:
Real-Ear Measurements (REMs)

 

[Mr. Haelscheir]

You might look at what hearing aid professionals call:
Real-Ear Measurements (REMs)

My use cases for the two types of microphone mounts include:

1. Blocked canal measurements: Custom-molded microphone plug for consistent fit and sufficient noise isolation for high-SNR headphone distortion measurements.
2. Custom real-ear measurements: For better translation between speaker HRTFs and headphone compensation EQs. The existing implementations I am aware of have the preamp hung from the ear and would not accommodate personalized headphone measurements. I am hoping to devise something more compact and secure for head movements, though I still need to confirm how the probe mic is insulated from sounds impinging on parts other than the tip of the tube. As such, I would have to calibrate my own transducers and electronics, if such calibration is even worthwhile.

As for what I am trying to get out of this?

1. The blocked canal measurements are to have personalized headphone distortion and frequency response measurements (particularly of flagships) with my own real human head as the coupler and influencing headphone seating, yielding better results than the miniDSP EARS without having to dish out on an expensive industry standard coupler that will not perfectly match my own ears. For this I would likewise like accurate SPL levels.
2. I am enjoying head-tracked binaural simulation per my findings in https://www.head-fi.org/threads/rec...-virtualization.890719/page-121#post-18027627 (post #1,812), but still feel that the stereo image comes from no further than my original 1.5 m distance HRTF measurements, whereby my hope is that longer-distance measurements will capture the ITD and phase relations needed for more convincing imaging of distance. I want to recreate the concert hall experience at home where I can't even regularly use my studio monitors, or be able to take an outdoor or gym measurement in order to binaurally simulate the stereo width of a 16 meter wide orchestra for which your typical audiophile is not going to be able to set up an equivalently wide listening room. It is an experiment in the capabilities of sound reproduction when one feels that going to the orchestra almost once a week doesn't satisfy me, or one does not want to have to wait a year or few for that amazing piece to be programmed again.