thersites
Advocate
  • 46
  • 110
  • 0
Registered since

Mar 24, 2017

directional oddity with noise having a short attack time

Hi,

 

In general the Hearphones are quite remarkable, but I've noticed a problem with relatively high-frequency noises having a short attack time and a relatively strong pulse, such as when silverware makes contact with a ceramic plate: the noise will be reproduced in the correct ear bud (i.e. the one closest to the noise), but then very shortly thereafter it will be reproduced in the *other* earbud, which is confusing and very noticeable.

 

Has anyone else experienced this? Surely this is a bug in the audio processing algorithm?

11 REPLIES 11
ST - Hearphones
Community Manager - Retired
  • 0
  • 2081
  • 160
Registered since

Dec 23, 2016

Re: directional oddity with noise having a short attack time

Hi thersites,

 

Thank you for telling us about what you're experiencing. We would like to understand more.

 

Please tell us about your settings and environment when you notice this.

  • World Volume
  • Directivity
  • Bass/Treble
  • Are you in a noisy environment or quiet?
  • Are you hearing the sound louder in the closest earbud to the source or the earbud on the other side?

Thank you,

ST

 

thersites
Advocate
  • 46
  • 110
  • 0
Registered since

Mar 24, 2017

Re: directional oddity with noise having a short attack time

Hi,

 

I've experienced this in the group conversation mode, with World Volume set to 20, and bass/treble to 0. Relatively noisy environment, but I doubt that that changes the behavior. Most easily reproduceable with focused or front directivity, but also with "everywhere." Earbud balance is 0.

 

The test sound I'm using is that produced by lightly hitting the thin side of a spoon against a ceramic plate. If the plate is in front of me to my left, what seems to happen is that the sound reproduced first in the slightly more distant ear bud, and then significantly more loudly in the earbud closest to the source of the noise. If I move the plate to my right, the opposite seems to be true: the sound is reproduced first in the right earbud, but then more loudly in the left earbud.

 

What is most disconcerting is that the sound appears to move from one earbud to the other, with a slight delay.

ST - Hearphones
Community Manager - Retired
  • 0
  • 2081
  • 160
Registered since

Dec 23, 2016

Re: directional oddity with noise having a short attack time

Hi thersites,

 

Thanks for providing more information. I can replicate this: hear it in one ear, and then the other ear 

I haven't noticed the flip in the direction. 

 

I need a little more time with this and I'll be back when I have more information.

 

ST

 

thersites
Advocate
  • 46
  • 110
  • 0
Registered since

Mar 24, 2017

Re: directional oddity with noise having a short attack time

Oh good. It is definitely odd and disconcerting as it isn't natural. I drove someone crazy yesterday at a café as I was experimenting with the sound!

ST - Hearphones
Community Manager - Retired
  • 0
  • 2081
  • 160
Registered since

Dec 23, 2016

Re: directional oddity with noise having a short attack time

Hi thersites,

 

You wrote: " I drove someone crazy yesterday at a café as I was experimenting with the sound! "

 

I shared your comment with a friend tonight. She wondered if the someone in your story was a companion, or a stranger at a different table.

 

Back to your question. This could explain your experiences.

 

There is an impulse noise reduction algorithm that runs independently in each ear.  Its purpose is to help avoid discomfort when an impulse (such as silverware striking ceramic) reaches one of the earbuds.  It does this by looking for signals with a fast attack time that are above some loudness threshold.  When such a signal is detected, the gain is temporarily reduced over the duration of the impulse (which by definition, is very short). 

 

An acoustic impulse may sound louder at one ear than the other if the impulse originates off to one side of the head.  The brain uses this level difference (along with a time difference) to judge direction.   In such a scenario, the impulse noise reduction algorithm may be triggered on the ear closer to the source, and not at the other ear, due to the defined loudness threshold.   This will cause the impulse to sound louder in the ear that’s farther from the source.   The relative prominence of the impulse in the ‘wrong’ ear may be producing the confusing and noticeable sensation you described.

 

What do you think?

ST

 

 

thersites
Advocate
  • 46
  • 110
  • 0
Registered since

Mar 24, 2017

Re: directional oddity with noise having a short attack time

I was bothering a stranger at another table Man Wink

 

I'm impressed by your knowledge of the underpinnings of the Hearphones. I find your explanation of the phenomenon in terms of the impulse noise reduction algorithm fairly persuasive.  But I find this behavior unexpected for the following reasons:

 

  1. The noise of silverware striking ceramic isn't *that* loud -- I'm just lightly tapping the silverwear against the ceramic plate (which is close to me, so the distance difference is minimal).
  2. Actually, the impulse seems to sound louder in the ear *closest* to the source -- but after a short delay. So the sound could be blocked by the earbud nearest the sound, but then, for some reason, after the sound has been suppressed by the algorithm, the sound is then replayed without suppression?!
  3. The effect seems to disappear if I play music over bluetooth (and it doesn't appear to be because the music masks the sound of the effect).
  4. My test hearing aids have a similar impulse noise reduction algorithm, and they don't behave this way.

 

Can you please report this to your contacts at Bose so that they can investigate?

 

 

ST - Hearphones
Community Manager - Retired
  • 0
  • 2081
  • 160
Registered since

Dec 23, 2016

Re: directional oddity with noise having a short attack time

Hi thersites,

 

thersites wrote:

I'm impressed by your knowledge of the underpinnings of the Hearphones.

 

I have the incredible fortune to be working with an exceptional team. We have been talking about your observations and questions.

 

More to come...

thersites
Advocate
  • 46
  • 110
  • 0
Registered since

Mar 24, 2017

Re: directional oddity with noise having a short attack time

Great. I hope the impulse noise suppression can be improved (if that is indeed the cause). Since meals are an obvious place for group conversation mode, the problem with short attack noises (silverware against ceramic plates) recurs frequently. Speaking of modes, it would be great to be able to associate them with locations (via geofencing) and have them switch automatically if desired.

 

By the way, I have found another test case. Try holding a keychain holding several keys directly in front of yourself and shake them.

ST - Hearphones
Community Manager - Retired
  • 0
  • 2081
  • 160
Registered since

Dec 23, 2016

Re: directional oddity with noise having a short attack time

Hi thersites,

 


@thersites wrote:

I was bothering a stranger at another table Man Wink


I'll pass that along.

 


I'm impressed by your knowledge of the underpinnings of the Hearphones.


I have the incredible fortune to be working with an exceptional team. We have been talking about your observations and questions.

 


 I find your explanation of the phenomenon in terms of the impulse noise reduction algorithm fairly persuasive.  But I find this behavior unexpected for the following reasons:

 

  1. The noise of silverware striking ceramic isn't *that* loud -- I'm just lightly tapping the silverwear against the ceramic plate (which is close to me, so the distance difference is minimal).

As you know, an impulse noise reduction algorithm factors in more than just volume. A more detailed discussion on topic is beyond the scope of this forum.


  • Actually, the impulse seems to sound louder in the ear *closest* to the source -- but after a short delay. So the sound could be blocked by the earbud nearest the sound, but then, for some reason, after the sound has been suppressed by the algorithm, the sound is then replayed without suppression?!

We've considered software and hardware issues that could cause asymmetry but these would degrade performance and symmetry under normal conditions as well. You would have noticed.

 

It is more likely that the environment where you experienced the issue is asymmetrical – there could be many factors in the surrounding environment that could make the loudness of the impulse vary as the source is moved from left to right, and which would influence the prevalence of the behavior of the impulse response algorithm.


  • The effect seems to disappear if I play music over bluetooth (and it doesn't appear to be because the music masks the sound of the effect).

It sounds as though the issue is being masked by music.


  • My test hearing aids have a similar impulse noise reduction algorithm, and they don't behave this way.

Hearphones are not hearing aids. Hearphones are conversation-enhancing headphones that are specially designed to help you hear in noisy environments.


Can you please report this to your contacts at Bose so that they can investigate?


Done.


I hope the impulse noise suppression can be improved (if that is indeed the cause). Since meals are an obvious place for group conversation mode, the problem with short attack noises (silverware against ceramic plates) recurs frequently.

 


Our team is working hard to improve all of our signal processing algorithms to maximize the benefits while minimizing artifacts. I can't promise if or when we can improve this particular issue. 
 


Speaking of modes, it would be great to be able to associate them with locations (via geofencing) and have them switch automatically if desired.


That's an interesting idea. Please consider posting that in Feature Requests for more visibility in the community.
 


By the way, I have found another test case. Try holding a keychain holding several keys directly in front of yourself and shake them.


Other than driving someone crazy (sorry, that was irresistable), I didn't experience anything unexpected.

 

We're working on improving Hearphones. Thank you for your observations, insights and suggestions.

 

ST