Sounds like a no-compromise approach to what Creative has already attempted to implement in the Super X-Fi
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Measuring HRTF for headphone use
- Thread starter pkane
- Start date
Sounds like a no-compromise approach to what Creative has already attempted to implement in the Super X-Fi
svart-hvitt
Major Contributor
- Joined
- Aug 31, 2017
- Messages
- 2,375
- Likes
- 1,253
I didn’t know Super X-fi. Thanks for heads up.
It’s probably same same but different. Let’s see
svart-hvitt
Major Contributor
- Joined
- Aug 31, 2017
- Messages
- 2,375
- Likes
- 1,253
Genelec’s R&D director Aki Mäkivirta held an AES speech on the new Aural ID in York today. From ca. 7:22:00 in this video:
^ the link seems to be broken...?
svart-hvitt
Major Contributor
- Joined
- Aug 31, 2017
- Messages
- 2,375
- Likes
- 1,253
Yes, it was there for a while an now it’s taken away. I don’t know what’s happened
Did he maybe reveal too much and Genelec took the video out? They seem to be quite secretive, for example I once asked them what DAC chips they use in their speakers and apparently it's classified...
They seem to be quite secretive, for example I once asked them what DAC chips they use in their speakers and apparently it's classified...
svart-hvitt
Major Contributor
- Joined
- Aug 31, 2017
- Messages
- 2,375
- Likes
- 1,253
Did he maybe reveal too much and Genelec took the video out?
No, it was not a Genelec only video. It was a video full of AES presentations, many hours long. Genelec even posted the video on their Facebook page so Genelec was probably happy with the content.
The Genelec presentation was 20 minutes long, including Q&A. Nothing revolutionary, but they stated clearly this is not an audiophile product.
svart-hvitt
Major Contributor
- Joined
- Aug 31, 2017
- Messages
- 2,375
- Likes
- 1,253
Did he maybe reveal too much and Genelec took the video out?
PS: @soundwave76 , on your remark regarding classified DAC chips in Genelec (or other active speakers): I think @amirm ’s measurements have told us long ago that there’s no need to obsess about neither DACs nor amplifiers if the designer is competent. Unwanted byproducts of DACs and amps are of such a low order compared to what happens in driver motors and in the room interaction. This is not easy for audiophiles to accept, who want to know all the details. In my experience, which is not audio related, what happens when you give away minor details is like opening Pandora’s box: People will start asking for ever more details, and will not see the wood for the trees.
In a narrow rational sense, more information is better than less information. In a more realistic rational sense, it’s better to have less details. The story of Ulysses made this point long, long ago.
I agree and do NOT obsess / care that much about DACs in Genelecs
Aural ID should be a step towards more meaningful use of headphones for monitoring (though the word "meaningful" maybe is a bit misused these weeks). We're helping to build a standardised railroad, and hope other will contribute better locomotives.
"Dumb" headphones exclude the influential external ear from our hearing, so they break the link to natural listening we have personally acquired over a lifetime. The outer ear together with movement is what provides us the wonderful feature, spherical hearing, which we constantly use, regardless if we work in stereo or immersive.
Considering stereo, for example, each of us receive direct sound in the 60 degree angle in a distinct way, like we receive cross-talk and room reflections. By describing how sound from all different angles is modified during normal arrival, considering You personally, Aural ID enables a rendering engine to offer natural (or supernatural) hearing, even when using cans. Important sources like human voice are difficult to level, pan and eq on normal headphones, because midrange translates randomly between people when using them. What you hear can be quite different from what the other person hears, even when you are using the same headphones.
Immersive is an obvious benefit of personal, spherical listening, with any number of discrete direct sound sources for the renderer to cope with, ideally also movements and room reflections.
Aural ID provides the foundation for headphone rendering in stereo as well as immersive. More documentation about the precision of the technique is required and will be available.
"Dumb" headphones exclude the influential external ear from our hearing, so they break the link to natural listening we have personally acquired over a lifetime. The outer ear together with movement is what provides us the wonderful feature, spherical hearing, which we constantly use, regardless if we work in stereo or immersive.
Considering stereo, for example, each of us receive direct sound in the 60 degree angle in a distinct way, like we receive cross-talk and room reflections. By describing how sound from all different angles is modified during normal arrival, considering You personally, Aural ID enables a rendering engine to offer natural (or supernatural) hearing, even when using cans. Important sources like human voice are difficult to level, pan and eq on normal headphones, because midrange translates randomly between people when using them. What you hear can be quite different from what the other person hears, even when you are using the same headphones.
Immersive is an obvious benefit of personal, spherical listening, with any number of discrete direct sound sources for the renderer to cope with, ideally also movements and room reflections.
Aural ID provides the foundation for headphone rendering in stereo as well as immersive. More documentation about the precision of the technique is required and will be available.
svart-hvitt
Major Contributor
- Joined
- Aug 31, 2017
- Messages
- 2,375
- Likes
- 1,253
More on Aural ID by R&D director Aki Mäkivirta of Genelec:
Q: «How does this compare to TrueFi by Sonarworks?»
A: «TrueFi by Sonarworks is an application that contains a headphone equalizer based on the average response measured in certain brand and type of headphones. This can reduce coloration in headphone listening. They also advertise listener-specific equalization or sound colour adjustment as well. TrueFi does not do anything like Genelec Aural ID. TrueFi is more directed towards consumers.
Aural ID is a reliable and accurate method of obtaining your personal Head-Related Impulse Responses (HRIR, in time domain) for hundreds of different orientations of arrival for audio. This information is also referred to as the Head-Related Transfer Function (HRTF, in frequency domain). HRIR and HRTF both explain how sound changes when it travels to your ears from a certain direction. The main changes have to do with amplification and attenuation of certain frequencies depending on the direction of arrival, and the timing and phase relationship related changes in audio and between the same audio signal arriving at your two ears. These combine to create your sensation of the direction for the audio signal. Aural ID contains this information in a file and is specific to you individually.
Who are those that could use Aural ID HRTF information at this time? All those who have information about the direction of arrival for sound. Examples of these potential users are Virtual Reality and Augmented Reality systems, game engines calculating the audio presentations dynamically as part of the gaming process, and researchers working with 3D audio. Ultimately, we all can benefit, once the method of presenting immersive audio develop to include information about the direction of sound arrival and the processing needed to plug in the personal HRTF information to obtain the correct presentation over headphones. This will elevate headphone listening to be more reliable.
Genelec HRTF measurement method is aimed at audio professionals, opening new possibilities to the whole audio industry. It does this in a fully personal level, maintaining complete accuracy and full personal detail, with robustness and convenience needed to bring the HRTF to all professional applications.
Best regards,
Aki».
Source: Facebook
Q: «How does this compare to TrueFi by Sonarworks?»
A: «TrueFi by Sonarworks is an application that contains a headphone equalizer based on the average response measured in certain brand and type of headphones. This can reduce coloration in headphone listening. They also advertise listener-specific equalization or sound colour adjustment as well. TrueFi does not do anything like Genelec Aural ID. TrueFi is more directed towards consumers.
Aural ID is a reliable and accurate method of obtaining your personal Head-Related Impulse Responses (HRIR, in time domain) for hundreds of different orientations of arrival for audio. This information is also referred to as the Head-Related Transfer Function (HRTF, in frequency domain). HRIR and HRTF both explain how sound changes when it travels to your ears from a certain direction. The main changes have to do with amplification and attenuation of certain frequencies depending on the direction of arrival, and the timing and phase relationship related changes in audio and between the same audio signal arriving at your two ears. These combine to create your sensation of the direction for the audio signal. Aural ID contains this information in a file and is specific to you individually.
Who are those that could use Aural ID HRTF information at this time? All those who have information about the direction of arrival for sound. Examples of these potential users are Virtual Reality and Augmented Reality systems, game engines calculating the audio presentations dynamically as part of the gaming process, and researchers working with 3D audio. Ultimately, we all can benefit, once the method of presenting immersive audio develop to include information about the direction of sound arrival and the processing needed to plug in the personal HRTF information to obtain the correct presentation over headphones. This will elevate headphone listening to be more reliable.
Genelec HRTF measurement method is aimed at audio professionals, opening new possibilities to the whole audio industry. It does this in a fully personal level, maintaining complete accuracy and full personal detail, with robustness and convenience needed to bring the HRTF to all professional applications.
Best regards,
Aki».
Source: Facebook
There's chicken and egg to this, because Aural ID won't show its full potential unless used in combination with a capable SOFA renderer, and they are not public. Many of the early users are therefore from academia, but here's a review and a recent blog post with more details,
https://issuu.com/bizmediauk/docs/am...august_2019/32
https://www.genelec.com/blog/how-doe...ening-part-two
https://issuu.com/bizmediauk/docs/am...august_2019/32
https://www.genelec.com/blog/how-doe...ening-part-two
Hipper
Addicted to Fun and Learning
Firstly thanks to everyone who has contributed to this thread about trying to correct the frequency response of headphones - I've found it very interesting.
I have some questions please/.
1. Head Related Transfer Function (HRTF) refers to a range of bodily influences on sound as it approaches and enters the ears. The Pinnae and other bits of the ear, and ear canal have been mentioned along at times with the torso (not sure this is included in HRTF), but no-one has really discussed the influence of the head shape and things like the nose and hair.
Along with the torso, these latter parts cause an effect when listening to speakers but are bypassed with headphones and will not be included with in ear microphone measurements. Our brains are used to these effects so surely they will be missed. How significant are they?
I do understand solderdude's comments that there's not much point to striving for measurement accuracy in this area because of all the pitfalls but for those that are trying to do this, how are you dealing with these matters? Or are you ignoring them!
2. I have larger then average ears and until I bought the HD800 (or a lot earlier, the Jecklin Floats) my pinnae were always deformed when using headphones. This must surely be a factor in how I perceive the sound from headphones. Again I don't know if the effect is significant but it is clear from speaker listening that deforming the pinnae can considerably alter the perceived sound.
3. With speakers we get interaural crosstalk - the left ear hears the sound from the right speaker as well as from the left. Of course we don't get this with headphones and some thought that was detrimental. 'Crossfeed processors' were developed for headphone users (e.g. with the Headroom Blockhead and a Meier amp which I've used) to add the missing sounds but I never really found it did much for me. The fact that most high end headphone amps don't bother with this suggests to me it's not that important.
Conversely, Ambiophononics is the idea that interaural crosstalk from speakers is bad for the stereo image and either physically stopping it with absorption or using a processor to counteract it is said to improve things. On top of this we have Toole saying that the phantom image from speakers is weakened at around 2kHz (by 6-10db it appears) because of comb filtering caused by this interaural crosstalk, but that side wall reflections can partially help fill this dip in.
Which is the best solution - stopping interaural crosstalk on speakers or adding it to headphones?
I have some questions please/.
1. Head Related Transfer Function (HRTF) refers to a range of bodily influences on sound as it approaches and enters the ears. The Pinnae and other bits of the ear, and ear canal have been mentioned along at times with the torso (not sure this is included in HRTF), but no-one has really discussed the influence of the head shape and things like the nose and hair.
Along with the torso, these latter parts cause an effect when listening to speakers but are bypassed with headphones and will not be included with in ear microphone measurements. Our brains are used to these effects so surely they will be missed. How significant are they?
I do understand solderdude's comments that there's not much point to striving for measurement accuracy in this area because of all the pitfalls but for those that are trying to do this, how are you dealing with these matters? Or are you ignoring them!
2. I have larger then average ears and until I bought the HD800 (or a lot earlier, the Jecklin Floats) my pinnae were always deformed when using headphones. This must surely be a factor in how I perceive the sound from headphones. Again I don't know if the effect is significant but it is clear from speaker listening that deforming the pinnae can considerably alter the perceived sound.
3. With speakers we get interaural crosstalk - the left ear hears the sound from the right speaker as well as from the left. Of course we don't get this with headphones and some thought that was detrimental. 'Crossfeed processors' were developed for headphone users (e.g. with the Headroom Blockhead and a Meier amp which I've used) to add the missing sounds but I never really found it did much for me. The fact that most high end headphone amps don't bother with this suggests to me it's not that important.
Conversely, Ambiophononics is the idea that interaural crosstalk from speakers is bad for the stereo image and either physically stopping it with absorption or using a processor to counteract it is said to improve things. On top of this we have Toole saying that the phantom image from speakers is weakened at around 2kHz (by 6-10db it appears) because of comb filtering caused by this interaural crosstalk, but that side wall reflections can partially help fill this dip in.
Which is the best solution - stopping interaural crosstalk on speakers or adding it to headphones?
- Joined
- May 24, 2016
- Messages
- 643
- Likes
- 2,408
On a somewhat related note, folks may be interested in this open source project: https://github.com/jaakkopasanen/Impulcifer
PS. love those https://www.soundprofessionals.com/cgi-bin/gold/item/MS-TFB-2 binaural mics. There is an XLR version for use with 48v phantom power if you are using a stereo mic pre like https://www.soundonsound.com/reviews/art-pro-mpa2 I recorded a song through my NAD Viso HP50 headphones and compared to the original digital track and surprised at how close the recording sounds to the original.
PS. love those https://www.soundprofessionals.com/cgi-bin/gold/item/MS-TFB-2 binaural mics. There is an XLR version for use with 48v phantom power if you are using a stereo mic pre like https://www.soundonsound.com/reviews/art-pro-mpa2 I recorded a song through my NAD Viso HP50 headphones and compared to the original digital track and surprised at how close the recording sounds to the original.
Thanks for the link - I have those binaural mics and have been intending to try to use them as part of an 'ambiophonics over headphones' experiment. Impulcifier isn't quite exactly that - it seems more like the Realizer in that it's trying to recreate real speakers in a real room - but my idea is a simpler subset and so the code should be able to be adapted.
My idea is that the significant difference between basic headphone listening and ambiophonics is that in ambio you have HRIR in play while in headphones you don't. (you also have room ambience, but those are likely secondary). It seems reasonable to think that if you use binaural mics to record a speaker coming from directly in front of you (ideally FIR eq'd to be flat in freq and phase), then that should directly provide the target transfer function for headphone EQ. Then just measure the headphone response with the same mics and create the correction filter. My plate is pretty full, but the existence of the Impulcifier project seems to indicate that the basic idea works well enough to try.
I did read up on HRIR processing a bit a while back, and looked at some of the archives of recorded responses. I was surprised at how short they were - either 256 or 512 samples in many cases.
To re-iterate my point from above, IMHO the idea of adding crosstalk to try to remedy the 'in the head' aspect of headphones is the wrong approach. It seems to me that the critical aspect to introduce is HRIR, since if you get that right you basically get the ideal version of what the ambiophonics guys are trying to do with their crosstalk cancellation.
If I actually get a chance to play with my binaural mics and try this, I'll definitely update the thread.
What we call cross-talk is a natural element of hearing individual sources, and one of the features enabling us to localise. Question is, if it’s always a benign mechanism, e.g. considering reproduction of composite content?
For those of us going to AES in NYC next month, David Griesinger will be playing excellent binaural recordings using ultra-nearfield monitoring with cross-talk cancellation; which is an elegant way of separating humanisation in recording from personalisation in reproduction. I suggest you go listen and make up your own mind.
For those of us going to AES in NYC next month, David Griesinger will be playing excellent binaural recordings using ultra-nearfield monitoring with cross-talk cancellation; which is an elegant way of separating humanisation in recording from personalisation in reproduction. I suggest you go listen and make up your own mind.
Long time no see folks! Was able to get my hands on the SP-TFB-2 binaural mics. They seem pretty nice and makes it way easier to do measurements in REW. Might post a graph of my recent headphone I purchased (Status CB-1) once I rewrap my head around the old post mentioned by @solderdude and @pkane
Might post a graph of my recent headphone I purchased (Status CB-1) once I rewrap my head around the old post mentioned by @solderdude and @pkane Similar threads
- Poll
