Skip to main content

Topic: using MathAudio for car audio correction (convert playlist using DSP) (Read 238 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.
using MathAudio for car audio correction (convert playlist using DSP)
I've been thinking about using MathAudio for in-car sound correction, as most car audio installations suffer from low quality speakers and severe "room" coloration.

Suggested setup:
  • laptop with Foobar and MathAudio plugin
  • audio interface for output and input
  • measurement microphone

Suggested procedure:
  • if possible, time-align speaker/subwoofer setup for the preferred listening-position (my car stereo has this possibility, improves the stereo image and sound quality greatly!)
  • set the EQ of the car stereo to neutral, and disable other sound 'enhancement' options (they only ruin the sound)
  • connect the laptop soundcard output to the external input of the car stereo (in my case through a mini-jack connection)
  • make sure the input doesn't clip the audio, so lower the soundcard output level if needed
  • run the MathAudio measurement a couple of times with the measurement microphone at the preferred location(s)
  • verify the correction by playing a couple of audio files from laptop
  • name and save the measurement
  • switch to the Foobar 'convert' menu
  • open the 'Processing' menu
  • select 'MathAudio Room EQ' and make sure the correct measurement is loaded
  • batch-convert a playlist to the preferred file format (WAV in my case)
  • copy the playlist to a USB stick/disk and connect to the car stereo
  • play and enjoy!

As I see it, batch-converting audiofiles using the Mathaudio correction curve as DSP-plugin wil 'embed' the correction into the audio files, thus creating an "room corrected" playlist for in your car.

In my case, the car stereo can play 24-bit 48kHz WAV-files so I'm thinking about adding an additional DSP-plugin first (SOX), to resample to 48kHz before the EQ-correction takes place (as 44.1kHz gets upsampled to 48kHz anyway through the messy DSP-chip of my car stereo. I rather have this upsampling done through SOX).

Furthermore, the 24-bit depth and 48kHz samplerate should offer some additional resolution for the EQ processing (at least in theory)

Any thoughts on my ideas?

I will try this out in the coming days and report back about my experiences!  8)

my setup:
  • laptop
  • iFi DSD iNano D/A convertor (for playing the measurement sweep sound)
  • Roland UA-101 A/D convertor with phantom power (for connecting the measurement microphone)
  • Behringer ECM8000 measurement microphone
  • Foobar with MathAudio Room EQ and calibration file for the microphone
  • Last Edit: 31 December, 2016, 01:29:46 PM by filthyaudiophile