Headphones calibration with added hearing lose compensation

Comments

11 comments

  • Official comment
    Avatar
    Kārlis Butins

    Hi,

    It would be beneficial if you perform the (audiometry) test on the actual headphones that you would be sending for calibration, also include the test results for us, so we can make an educated calibration specifically for you.

    Thank you for your feedback and feature suggestion, we haven't received many of these requests as of yet, but we will definitely keep that in mind for future versions of our software.

    Comment actions Permalink
  • Avatar
    Eric Zang

    This is a great idea. I would also like to be able to have the option of applying audiometry results to the calibration.

    2
    Comment actions Permalink
  • Avatar
    Porangui McGrew

    Fantastic idea +1  

    Please include me with this too.

    2
    Comment actions Permalink
  • Avatar
    Randy Dillon

    +1 Yup hearing loss compensation would be awesome! Compensating for my hearing loss while mixing is something I've totally fudged (poorly) by adding an EQ on my master bus. Calibration seems the obvious best place to compensate for hearing loss.

    Suggested steps:

    The first simplest thing would be to add the ability to simply load a "personal" calibration file (don't need to call it "hearing loss") in addition to (i.e. NOT instead of) the existing calibration options (although if the "custom" reference curve could be made to load a file, or define an actual multi band curve, that would be a useful step. However that cannot be used at the same time as the predefined reference curves, and really this "personal" hearing loss calibration needs to be able to be summed to all other reference curves that might be applied [e.g. I want "flat" monitoring that includes compensaton for my hearing loss and headphones/speakers, and if I choose some of the other reference curves, or simulations, I still want my hearing loss compensation added in there as well])

    Next would be able to actually define and save this personal calibration curve, like a multi-band EQ.

    Finally, having a hearing test that can be performed after the normal calibration is engaged for the headphones/speakers would be most awesome. My self-test was done on my phones, and of course the test is colored by the phones themselves not having a flat response. I can now re-do such a test with my phones calibrated to the average (DT 770 Pro 80  ohm), but I still have to use an additional EQ plugin to compensate for it. 

    And of course the "personal" curve needs to be able to easily be disabled to easily audition for other listeners.

    0
    Comment actions Permalink
  • Avatar
    Sky Doherty

    +1 for a personal hearing eq curve.  Certainly, using standard and free hearing tests such as this would work, then Sonarworks could just create an EQ curve for correction.  https://hearingtest.online 

    0
    Comment actions Permalink
  • Avatar
    Troy Hall

    I have just this week seen an audiologist about my hearing loss, and amongst other things, I asked about how it's measured and whether I can use those measurements to configure EQ for each ear. I picked my audiologist because of their extensive experience with musicians, and learned quite a bit.

    While is is theoretically possible to create an EQ curve from the hearing test results, the devil is in the details. It turns out that for most people, applying 100% compensation will (at best) sound unbearably unnatural to the person with the hearing loss. After years of gradual declines, a sudden reversal will sound terrible. Second, the amount of compensation required in some frequency bands would introduce noticeable distortion in the output signal for most if not all EQs that listeners would have access to.

    In other words, even if you had an EQ that could adjust frequency bands without distortion, the result would still sound unnatural to the listener. More to the point, if you think you're going to get perfect hearing back by applying careful EQ, you are deluding yourself. However, it turns out that you can make things sound much better to yourself and this can help you make better mixing and sound design choices, but only if you give up the idea of "perfect".

    First, having a hearing loss curve may help you set *relative* correction levels which you can apply as a whole but to a smaller degree. For example, you can apply 25% of the correction curve, where the correction at each frequency is 25% of the total correction for that frequency.

    Second, you can adjust your correction curve so that any corrections that are way out there, like a 40 db correction at 8k Hz, are reduced to a level that won't distort.

    Third, if you accept that you won't ever hear things in the high frequency range correctly, and understand that you'd probably not make any significant changes up there even if you *could* hear them, you can just leave that part of the EQ alone, and let a master engineer work on the "air" or whatever else would make the recording sound better to people that hear better than you do.

    I've used the Aumeo device to apply some corrections, and it's kinda good/kinda OK. It definitely has helped me make adjustments that translate a lot better to different listening devices and contexts. It has helped my mixes for sure. It's definitely not perfect and doesn't sound natural to me.

    However, the problem with using separate devices and plug-ins is that costs more, introduces complexity in a setup, uses more CPU, adds more latency, and it seems like a hack when you already have software that makes the same kinds of adjustments you are after, and with better technology for doing it.

    So, Sonarworks, please please please consider adding a feature that would let a user (or maybe a user's audiologist) configure a hearing loss profile, and then enable that hearing profile to be applied in part (e.g. 0-100%), and include features aimed at reducing distortion like adding an option of limiting the amount of correction applied to any given frequency band.

    I am sure a ton of musicians and producers have hearing loss, and either don't know it, or don't want to confirm it, or don't have the financial resources to do something about it. Adding a feature for compensating for hearing loss will help remove the stigma around it, and also help the world produce better sounding music.

    Thanks!

    0
    Comment actions Permalink
  • Avatar
    Tzechiu Lei

    I too believe this would be a GREAT feature, not only for hearing loss but also for tinnitus.

    Furthermore, since we are not robots, each and every ear arguably has a unique frequency response.

    0
    Comment actions Permalink
  • Avatar
    Matteo Cremolini

    +1

    Any news?

     

    Thanks!

    0
    Comment actions Permalink
  • Avatar
    Kārlis Stenders (Edited )

    Hi All,

     

    Thanks for staying engaged and sharing your thoughts on this! 

     

    Yes, this is definitely something that has not been addressed in the pro audio production software market, at least not successfully on a large scale. There is, in fact, news on this subject - we have released our new SoundID app in the consumer market for music listening, which does feature a hearing test that we have been working on over the last two years. It is not a medical-grade hearing test, but it will make a huge difference to the SoundID profile that is generated, in combination with headphone calibration (based on Reference, of course) and the preference test based on machine-learning algorithms, which we've also been developing. 

     

    So currently this hearing test has debuted in SoundID - you can try it for free and see what it's all about here. The development process on this has pushed us on a steep learning curve, so there is definitely potential for taking advantage of it in Reference too. It is, however, way too early to make any promises for actual development concepts or timelines since any sort of individual hearing compensation would have to be some sort of variation of a medical-grade hearing test, or close to it.

     

    So, despite the fact that this is currently not on the roadmap for Reference, the lessons learned and data gathered during the hearing test development in SoundID has definitely put us on a path where this would be possible in the future for Reference too!

    0
    Comment actions Permalink
  • Avatar
    Jeremy

    Yes, please make it possible to import the hearing test results into Sound ID Reference for creating a compensation curve (with 0-100% adjustment). Seems like a no-brainer considering what your software is designed to achieve.

    0
    Comment actions Permalink
  • Avatar
    Andrew Macaulay

    Yes please - I've got quite a significant loss from 1kHz up on one ear (standing too close to the speakers when playing keyboards live, mostly one way around) so using your learnings from SoundID and how to apply hearing compensation without the issues of over-compensating what has, for many, become a "learned" listening experience AND ensuring that it also does not cause any distortion from the compensation would be a really great option - as long as it remains flexible in terms of "how much" to be applied (allowing for the way you tend to learn to hear with these changes over time) and allowing for standard audiology reports as input as well as allowing for more accurate models for pro-engineers.

    0
    Comment actions Permalink

Please sign in to leave a comment.