[jude]

Impressive presentation and so well delivered!

Question about the ear canal modeling part: isnt it LESS accurate to smooth out the model for the ear canal? I realize that the averaging process necessarily smooths out the natural texture to generate the average ear canal in the first place, but is it not true that texture / bumpiness in the ear canal is itself a quality that we find in every ear canal?

Like comparing a smooth wall to one with bumps on it i guess. If we average ten bumpy walls we may get a smoother wall, but if it’s not bumpy anymore, does it still represent the essence of what the walls had in common?

Just spit balling here but I wonder if the ear canal model would be a better simulation if we went back post-averaging and artificially added some topology to the smoothened result at certain spots.

Thoughts?

@Shabda, thank you for the kind words! Yours is a very good question, and I think the answer to what you're asking can be found in An Average of the Human Ear Canal: Recovering Acoustical Properties via Shape Analysis.

The paper isn't very long, and I think it may take a read of it in its entirety to answer your question.

 

[Shabda]

@Shabda, thank you for the kind words! Yours is a very good question, and I think the answer to what you're asking can be found in An Average of the Human Ear Canal: Recovering Acoustical Properties via Shape Analysis.

The paper isn't very long, and I think it may take a read of it in its entirety to answer your question.

I really did have to read to the end to get my answer!

I first noticed that this research depends heavily on the validating power of Physical Impedance Measurements. No problem for me but just a note.

I got my answer in Figure 9, which presents good evidence that any natural bumpiness in a real ear is either not affecting the Physical Impedance of the ear canal or that their Average Ear Simulation Model adequately captures the acoustic properties of a real ear.

Figure 9 essentially shows that the “Average ear - simulation model” measures very similarly to the “Closest individual ear - simulation model”, which would have that bumpiness I seem to be concerned about

And the Average Ear Model measure very simularly to the Average Ear (real)

ipso facto bingo bango

Model good (at least for the caucasian population)

but they actually have some good evidence that variations in sex and age are minimal, so race may not a play a significant role. Nevertheless, you have to wonder!

Thoughts?

 

[dadracer2]

If you haven't seen it already, HBK posted my presentation from this year's HBK Product Physics Conference. While only 20 minutes long, it covers quite a lot of ground.

You can watch it here:

​

I thought it was going to be too techy or dry and boring but it was none of those things. In fact it is one of those situations where you wonder why no-one had done this before. In any event I wanted to ask given your background scene was filled with HD800 and HD800S if you have measured these with the new 5128 and then compared that to the measurements from the old system? This would be very interesting to see........please?