CHANGSTAR: Audiophile Headphone Reviews and Early 90s Style BBS

  • December 31, 2015, 11:08:21 AM
  • Welcome, Guest
Please login or register.

Login with username, password and session length
Advanced search  

Author Topic: LUT for Headphones?  (Read 719 times)

0 Members and 1 Guest are viewing this topic.

Synnöve

  • Swabbie
  • Brownie Points: +3/-0
  • Offline Offline
  • Posts: 9
  • A queer, feminist audiophilic cinephile.
LUT for Headphones?
« on: May 16, 2014, 08:00:21 AM »
If any of you operate in the nerdy consumer or professional digital video arenas, there exists what's called a 3D look up table (LUT). The idea is that any display won't be able to perfectly replicate a color space perfectly (REC709, used in blurays and HDTV, for example), so what one does is measure a large series of samples (2k to 9k) of color to determine how far from the mathematical representation of the color space the display actually is. Once this is done, a point cube with thousands of points is used for morphing the incoming content in to a signal that will play through the display with less errors.

Basically this is like the equalization of headphones. I understand equalization is done primarily by ear using cobbled together software, but why not follow the route of having profiles of headphones? Smyth does this with their realizer by having small microphones inserted in to the ear while samples are played through the headphones. This could certainly be done by enthusiasts for the effect of equalizing headphone output for an individual's ears.

The ideal system would mirror what is done in the printing or digital video industries:

-Build a source device for your headphone that can hold a profile.
-This profile is made using software that generates samples to be recorded via in ear mics and played using one's headphones.
-The profile will offset incoming signals so that outgoing signals, when played through the phone's to an individual's ears, will mirror a linear response.
-When combined, the music will get closer to sounding similar to a variety of observers.
Logged
The human brain is one of the lowest entropy structures known to exist. Given it's unfathomable links, connections, and operations, how can we have such brash confidence in our ability to accurate claim causation when it comes to auditory phenomena?

Solderdude

  • Grab the dScope Kowalski!
  • Able Bodied Sailor
  • Pirate
  • ***
  • Brownie Points: +206/-4
  • Offline Offline
  • Posts: 907
  • No can do skipper, the dScope was terminated
    • DIY-Audio-Heaven
Re: LUT for Headphones?
« Reply #1 on: May 16, 2014, 08:56:55 AM »
Alas this isn't really possible in a practical way.  :(
The thought is nice though and has crossed my mind long ago as well (and minds of others).

It is the measuring process that is the bottleneck here and IMO also the digital artefacts that will be the result of the 'correction'.  :-Z
The fear audiophiles even have for similar performing DACs sounding different will keep audiophiles (the intended victims of such expensive gear) away from a circuit with onboard [ AD conversion -> DSP -> DA conversion ] and adding that in their 'purist' chain not knowing what it will f-up and if the benefits outway the downsides for them.

The major practical problem is seen very clearly in Tyll's raw plots and this is done on a real dummy head !
Different positions will yield VERY different correction 'curves'.

The correction obtained from a mic, that most likely won't be in the correct position to begin with, will yield different results when a headphone is moved slightly.
What would be the right procedure, position and the end result ?

The realiser is something that looks at time domain, reflections and FR (but probably in a coarse way) and assumes the supplied STAX SR-207 is used.
Would like to hear it some day but won't buy it as I don't want/need/like to emulate speakers or surround and mostly listen to stereo signals.

AccudioR and similar programs already do something similar BUT can do so only from the headphone measured (which may differ from another one) on their rig with their correction curve which may or may not be accurate.

I am a FIRM believer in proper EQ and elaborate digital EQ is not amongst them as that introduces pre-ringing and postringing at audible frequencies.
« Last Edit: May 16, 2014, 10:12:05 AM by Solderdude »
Logged
Use your ears to enjoy music, not as an analyser.

Marvey

  • The Man For His Time And Place
  • Master
  • Pirate
  • *****
  • Brownie Points: +555/-33
  • Offline Offline
  • Posts: 6698
  • Captain Plankton and MOT: Eddie Current
Re: LUT for Headphones?
« Reply #2 on: May 16, 2014, 05:45:54 PM »
Impulse response for headphones. Process through DSP like in JRiver MC which has a filter for so. So far I have not been successful. Need to review documentation on it... Not enough time in the day.
Logged