Launch of myHRTF Stereo - New Lifelike Personalized 3D Sound for Headphones

[ObjectiveSubjectivist]

Thank you for your kind offer, but let me check our database first: which headphone model(s) are you interested in?

Grell OAE1

 

[myHRTF]

Grell OAE1

I hate to disappoint you but we have measured OAE1 and decided to not process and upload it since it shows frequency response anomalies that cannot be corrected by DSP. As a result, our filters may not work as intended and this could be blamed on us.
This is very unfortunate since OAE1 features an outstanding look and feel.

Do you have any other headphones at hand?

 

[ObjectiveSubjectivist]

I hate to disappoint you but we have measured OAE1 and decided to not process and upload it since it shows frequency response anomalies that cannot be corrected by DSP. As a result, our filters may not work as intended and this could be blamed on us.
This is very unfortunate since OAE1 features an outstanding look and feel.

Do you have any other headphones at hand?

Yes, a couple, he400se DMS Open Omega (diy), and LCD-2.

I'm mainly using oae1 and dms open omega because of comfort and light weight. Also I find OAE1 to be working well with my hrtf as many many other headphones are just too bright to my ears.
Not sure why

 

[myHRTF]

Yes, a couple, he400se DMS Open Omega (diy), and LCD-2.

I'm mainly using oae1 and dms open omega because of comfort and light weight. Also I find OAE1 to be working well with my hrtf as many many other headphones are just too bright to my ears.
Not sure why

Awesome, I recommend to try it with HE400se then. I understand your comment about headphones sounding too bright, but this should be solved with our filters, too. Moreover, with the proper ear shape setting, it should sound equally good independent from the headphone model.

Currently, we are working on the new version with less reverberation and I recommend that you test this new version. I just need to ask you for a bit more patience since we will need another 1-2 weeks from now.

 

[myHRTF]

I'm happy to announce that the new version is online now. It comes with several improvements and addresses most if not all points that were mentioned in the feedback. We now have both acoustics of a dry recording studio and a more live sounding livingroom in addition to the two concert halls, see details below.
Various new Hifiman models are now supported, too.

You may try it out for free under the following link, you only need your headphones and a ruler:
Personalized Filter Designer

PM me to get your favorite songs filtered offline for testing.

The shop will be closed for maintenance until mid of February.

Happy listening,
Philipp

Synthetic room acoustics
myHRTF Personalized Filters contain different synthetic room acoustics in order to adapt them to the use case and recording. All rooms feature the following characteristics:

• No reflections within the first 5 ms after arrival of the direct wave, with the exception of wanted reflections from your body as part of your HRTF. This avoids that reflections affect the HRTF.
• No strong reflections within the first 20 ms after arrival of the direct wave, in line with Recommendation ITU-R BS.1116-3. This allows to hear the room acoustics captured on the recording rather than the listening room.
• Reflections from favorable directions only, i.e. no reflections from above or behind. This avoids that reflections with little contribution to externalization and envelopment decrease imaging.
• No resonances / room modes (in line with Recommendation ITU-R BS.1116-3), for a dry and precise yet fullbodied bass frequency range.
• No flutter echoes (in line with Recommendation ITU-R BS.1116-3), for a smooth and clear treble range.

Recording Studio​

Recording studio (Studio) simulates a perfectly symmetric room with 42 m² and approx. 270 ms reverberation time (above 200 Hz), in line with Recommendation ITU-R BS.1116-3. Studio sounds very dry, the reverberation is hardly perceivable.

• Speaker distance 1 m: recommended for mixing and mastering in the nearfield, as well as gaming and virtual reality.
• Speaker distance 3 m: recommended for mixing and mastering in the farfield and for listening to dry stereo recordings that contain little reverberation. In addition, it is recommended for multichannel recordings and TV/movies.

Livingroom​

Not to be confused with an average livingroom, this is a large acoustically treated and perfectly symmetric room with approx. 480 ms reverberation time (above 200 Hz), with a speaker distance of 3 m. Livingroom sounds more live than Studio, the reverberation is perceivable in a pleasant way and contributes to externalization. It is recommended for stereo recordings that contain reverberation.

Concert and Recital Hall​

Stereo sound reproduction in typical listening rooms fails to create the sense of space and envelopment that can be experienced in concert halls. myHRTF Personalized Filters add this spaciousness and envelopment to the music and turn your headphones into a recital or concert hall by mouse click. They do so by adding reverberation with optimum reverberation times as recommended by experts such as Leo Beranek. Recital Hall features a reverberation time (above 200 Hz) of approx. 1.6 s, concert hall features approx. 1.9 s.

 

[Bugal1998]

@myHRTF I've been thinking about possible improvements to the HRTF customization process, as I'm not convinced I was able to get the settings correct for me (some of the measurements are challenging).

Instead of making input selections on a static page, could you enhance the process to be realtime and dynamic with test tones? For example, a signal of some sort is played (pink noise, a spoken voice, etc.) and the listener moves a slider to position the sound where it should be perceived (e.g. eye level and straight ahead, sounds at 45 or 90 degrees left and right, completely diffuse, etc., etc.); essentially a test tone and slider for each HRTF input.

If the final test tones are all performing as expected, then the listener can be confident their HRTF is optimized.

 

[myHRTF]

@myHRTF I've been thinking about possible improvements to the HRTF customization process, as I'm not convinced I was able to get the settings correct for me (some of the measurements are challenging).

Instead of making input selections on a static page, could you enhance the process to be realtime and dynamic with test tones? For example, a signal of some sort is played (pink noise, a spoken voice, etc.) and the listener moves a slider to position the sound where it should be perceived (e.g. eye level and straight ahead, sounds at 45 or 90 degrees left and right, completely diffuse, etc., etc.); essentially a test tone and slider for each HRTF input.

If the final test tones are all performing as expected, then the listener can be confident their HRTF is optimized.

Thank you for your feedback and ideas!
Since the first two measurements are rather easy and don't need to be so accurate, the last one (ear shape) is troublesome. For the time being, we will not get around the listening test in the last step, and for that pink noise or a voice would help maybe.
With the introduction of surround and immersive audio we will see more listening tests with directions as indicated by you.

Thank you again for your extensive feedback on the old version: you were spot on with your points!
I think that our new version addresses all of them, and as you have indicated, not all of them were related to the reflections. Looking forward to your feedback on the new version, especially with respect to the comparison with your speaker setup.

 

[myHRTF]

Our online shop contains the new version now and is open again. We will keep the launch price until market introduction of the surround extension which is planned for spring this year.

Looking forward to your feedback on this new version which is significantly improved and addresses all points in your valuable feedback (see details above).

You may try it out for free under the following link, you only need your headphones and a ruler (no registration required):
Personalized Filter Designer

PM me to get your favorite songs filtered offline for testing.

Happy listening,
Philipp

 

[Robbo99999]

Our online shop contains the new version now and is open again. We will keep the launch price until market introduction of the surround extension which is planned for spring this year.

Looking forward to your feedback on this new version which is significantly improved and addresses all points in your valuable feedback (see details above).

You may try it out for free under the following link, you only need your headphones and a ruler (no registration required):
Personalized Filter Designer

PM me to get your favorite songs filtered offline for testing.

Happy listening,
Philipp

How will the surround function work? It wouldn't be compatible with gaming dacs that currently mix 7.1 channels from a game into stereo headphones and therefore the dac already does the virtual surround processing - this is what I'm thinking as you don't want two lots of different virtual surround processing happening at the same time. Yeah, so how does your surround option work?

 

[myHRTF]

How will the surround function work? It wouldn't be compatible with gaming dacs that currently mix 7.1 channels from a game into stereo headphones and therefore the dac already does the virtual surround processing - this is what I'm thinking as you don't want two lots of different virtual surround processing happening at the same time. Yeah, so how does your surround option work?

Hi Robbo99999,
I guess you are referring to gaming on a Windows PC?
In this case I recommend to use Equalizer APO (EAPO) in combination with one of our configurations created specifically for EAPO (included in myHRTF Personalized Filter set): you assign EAPO to your gaming DAC and filter all sound going through it in realtime using our filters. I'm using SoundBlaster X4 this way and it works like a charm.

This is described here in more detail.

In addition, you need to switch off / bypass spatial processing from your gaming DAC to prevent double filtering.

 

[Robbo99999]

Hi Robbo99999,
I guess you are referring to gaming on a Windows PC?
In this case I recommend to use Equalizer APO (EAPO) in combination with one of our configurations created specifically for EAPO (included in myHRTF Personalized Filter set): you assign EAPO to your gaming DAC and filter all sound going through it in realtime using our filters. I'm using SoundBlaster X4 this way and it works like a charm.

This is described here in more detail.

In addition, you need to switch off / bypass spatial processing from your gaming DAC to prevent double filtering.

You'd keep your DAC in 7.1 mode would you and turn off the virtualisation in the gaming DAC. That way the game could output 7.1 and it would just be downmixed without HRTF involvement from the DAC software/hardware. Is that how a person would do it? In that case your personalised filter set (which you import into EqualiserAPO) has a different set of filters for each of the 7.1 channels to reflect the HRTF differences? So it's basically just applying the correct HRTF differences to each of the channels?

 

[myHRTF]

You'd keep your DAC in 7.1 mode would you and turn off the virtualisation in the gaming DAC. That way the game could output 7.1 and it would just be downmixed without HRTF involvement from the DAC software/hardware. Is that how a person would do it? In that case your personalised filter set (which you import into EqualiserAPO) has a different set of filters for each of the 7.1 channels to reflect the HRTF differences? So it's basically just applying the correct HRTF differences to each of the channels?

Correct, the game outputs 7.1 to the gaming DAC, and the spatial processing of this DAC is switched off / bypassed. EAPO is assigned to this DAC, it filters (renders binaurally) all 8 channels in realtime and outputs the result (binaurally rendered 7.1) at the left and right channels of the DAC.

Technically, it is the same that HeSuVi does, but with personalized binaural room impulse responses (BRIR) instead of generic ones, and without HeSuVi being required.

 

[Robbo99999]

Correct, the game outputs 7.1 to the gaming DAC, and the spatial processing of this DAC is switched off / bypassed. EAPO is assigned to this DAC, it filters (renders binaurally) all 8 channels in realtime and outputs the result (binaurally rendered 7.1) at the left and right channels of the DAC.

Technically, it is the same that HeSuVi does, but with personalized binaural room impulse responses (BRIR) instead of generic ones, and without HeSuVi being required.

(I've never used HeSuvi, it's a software processing add on to EqualiserAPO or works alongside it) No but more what I'm getting at with my questions, you're leaving EqualiserAPO unchanged? You're just providing personalised HRTF frequency response filters that are specific for each of the 7.1 channels and then just relying on the normal downmixing of the DAC (no HRTF effects applied by the DAC) from 7.1 to 2 channels. I'm just trying to understand & visualise how it all fits together as then it will give me an idea of theoretically how it might perform. If it does work like I describe, then I don't see anything negative in that, instead that's how I imagined it would function, and I think that could work - you still need a DAC that can be seen by the OS as a 7.1 channel device though if it works the way I describe it otherwise the game won't output 7.1 and the OS/EqualiserAPO won't be configured in 7.1 mode either. I do actually like the theory of this solution & I think it's got a good chance of working really well, but yeah if it works the way I think it works then you still need a 7.1 DAC.

 

[myHRTF]

you're leaving EqualiserAPO unchanged?

Yes, we just load our configuration file including the personalized BRIR.

You're just providing personalised HRTF frequency response filters that are specific for each of the 7.1 channels and then just relying on the normal downmixing of the DAC (no HRTF effects applied by the DAC) from 7.1 to 2 channels.

Exactly, except that the combination of binaural outputs for each of the 8 channels after binaural rendering (down-mixing) happens on your PC in EAPO.

you still need a DAC that can be seen by the OS as a 7.1 channel device

Correct, you need a multichannel audio interface with sufficient channels as seen by the OS. But it is not required that those channels are available as physical outputs at this interface, a stereo output is sufficient due to the down-mixing described above.

 

[Robbo99999]

Yes, we just load our configuration file including the personalized BRIR.


Exactly, except that the combination of binaural outputs for each of the 8 channels after binaural rendering (down-mixing) happens on your PC in EAPO.


Correct, you need a multichannel audio interface with sufficient channels as seen by the OS. But it is not required that those channels are available as physical outputs at this interface, a stereo output is sufficient due to the down-mixing described above.

For that middle point you make, does that mean you include an instruction in EAPO to downmix all the channels to stereo? I suppose that's not strictly necessary when you've got a 7.1 dac which has a stereo headphone output as the DAC would do the downmixing, but it doesn't hurt either way.

 

[myHRTF]

For that middle point you make, does that mean you include an instruction in EAPO to downmix all the channels to stereo?

Exactly, our filter sets come with EAPO configuration files (.txt files) that contain the necessary code. The signal path is 100% under our control this way, including automatic adaption to the sampling rate.

 

[Robbo99999]

Exactly, our filter sets come with EAPO configuration files (.txt files) that contain the necessary code. The signal path is 100% under our control this way, including automatic adaption to the sampling rate.

I'd be interested to try the virtual surround sound option of this for gaming when you've launched it. (Or before if you want my feedback on it. I think I've still got my personal anatomy measurements I did for this when I tried the 2 channel version of this on my HD800 headphones).

 

[myHRTF]

Fantastic thanks, I will be looking out for that

My apologies for the delay, the work on the room acoustics was way more complex than anticipated at first. With the launch of the new version, Arya Organic is supported, too.

 

[Joe Bloggs]

Hi! Just here to say that I have tried customizing on the myHRTF page and obtained my own set of customized BRIRs accordingly. I would give the results a solid 7/10 (for myself, results may obviously vary for everyone) and bonus points (not included) should be given for the easy customization process. This product is a great thing to try for anyone who has not yet obtained a good BRIR for themselves and I may link to this in my article on efofi.com in the future. The price is not for everyone but you get to try the effect online before buying. What would be even better would be if there's a facility to upload your own music for trialling before the purchase.

The in-front localization results were quite positive and at least on par with what I had been able to achieve with mic measurements in my own ear with my limited techniques.

Question to people here, what have your experience been with the tonality of the resultant filters? Do they go well with the earphones you selected? They have recently added a "neutral" "brand" of in-ear earphones you can get out of the purchased bundle on the conversion website, the idea being that if you're 100% satisfied with the tuning of your headphones / earphones before HRTF, you can select that instead of the actual brand / model. In these days of common squig.link measurement results and multiple tuning targets, this may be the way to go. I personally feel that the "neutral" results are a bit more coloured than to my liking, but of course the colouration itself helps with the front localization. I just feel that the front localization obtained with a "more neutral neutral" filter can still be good enough while being more pleasant tonally, whereas a "full-blown" filter as given now might make for a very out of your head presentation initially but which feels unnatural (at least without head tracking) and which collapses back in to a greater extent over time (again, without head tracking) compared to a more conservative spatialization in the first place.

 

[myHRTF]

Hi! Just here to say that I have tried customizing on the myHRTF page and obtained my own set of customized BRIRs accordingly. I would give the results a solid 7/10 (for myself, results may obviously vary for everyone) and bonus points (not included) should be given for the easy customization process. This product is a great thing to try for anyone who has not yet obtained a good BRIR for themselves and I may link to this in my article on efofi.com in the future. The price is not for everyone but you get to try the effect online before buying. What would be even better would be if there's a facility to upload your own music for trialling before the purchase.

The in-front localization results were quite positive and at least on par with what I had been able to achieve with mic measurements in my own ear with my limited techniques.

Question to people here, what have your experience been with the tonality of the resultant filters? Do they go well with the earphones you selected? They have recently added a "neutral" "brand" of in-ear earphones you can get out of the purchased bundle on the conversion website, the idea being that if you're 100% satisfied with the tuning of your headphones / earphones before HRTF, you can select that instead of the actual brand / model. In these days of common squig.link measurement results and multiple tuning targets, this may be the way to go. I personally feel that the "neutral" results are a bit more coloured than to my liking, but of course the colouration itself helps with the front localization. I just feel that the front localization obtained with a "more neutral neutral" filter can still be good enough while being more pleasant tonally, whereas a "full-blown" filter as given now might make for a very out of your head presentation initially but which feels unnatural (at least without head tracking) and which collapses back in to a greater extent over time (again, without head tracking) compared to a more conservative spatialization in the first place.

Thank you, Joe. Your feedback is greatly appreciated!