Dec 26, 2024 at 6:52 AM Thread Starter Post #1 of 1

gergo0217

New Head-Fier
Joined
Sep 28, 2015
Posts
1
Likes
10
Hello All,

I thought I'd start a thread here about a little project I've been working on with my Shure SE846 IEMs.

The project produced a phase correction Finite Impulse Response (FIR) filter that converts the phase response of the SE846s to a minimum phase without affecting the magnitude response at all. This can be applied to the playback chain with the SE846 on the end using a convolution DSP engine, similar to how room correction is done for loudspeakers. While the filter introduces some latency, it also creates a phase shift that is the (approximate) inverse of the SE846's non-minimum phase components. The net effect is a minimum phase response at the eardrum with the SE846's frequency response magnitude totally untouched, along with some sound quality improvements! Of course, the latter is subjective, but I'm hopeful that there will be some SE846 owners on this forum who can try it and confirm or deny. I would welcome any feedback to refine the method I developed.

All project details and the filter itself are available here:

https://github.com/GergoOrosz/IEM-Phase-Correction/tree/main

As an acoustician with expertise in various forms of active sound field control (including active noise cancelling and 3D audio), I started this pet project for three reasons:
  1. Pretty much every piece of gear I get, I tend to assess through measurements. The impulse responses I measured with my SE846s clearly showed that there is scope for improvement in the time domain.
  2. Nearly all objective sound quality assessments are done considering the frequency response magnitude alone. Assessments using various target curves are really popular these days, which is great to see, but I've heard claims that since most wearable audio devices have a minimum phase response anyway, the magnitude response should be sufficient to fully characterise the sound. This isn't true for multi-way IEMs (with a handful of exceptions). I wanted to see if I could improve the sound of my SE846 using DSP without altering the magnitude response.
  3. Have a way to be able to quickly switch between the original and minimum phase response to see how much of a sonic difference there is.

Some limitations may obviously apply.

I expect that most Balanced Armature-based 3+ way IEMs could benefit from digital phase response correction, and I'm hopeful that the SE846 is just the start of an exciting journey. I look forward to interesting conversations here – let me know what thoughts, questions, and comments you may have! Get in touch if you would like to know more.

Cheers
 

Users who are viewing this thread

Back
Top