docs: Something is missing in HOA-NFE explanation

[madskjeldgaard]

Hey everyone!

Thanks again for the amazing work here. I've been reading the HOA-NFE file today, and I've come to a realization as to why I am having a bit of trouble understanding this document.

It seems to me that in the explanation of the virtues of the HOA-NFE variant, there is something missing that might make it easier to understand, specifically in this part:

"It could be argued that the greatest innovation in Daniel's reframing of Ambisonics in a higher order context is the translation of the basic wave to a radius other than infinity.

If we'd like to think in a real world way, this is equivalent to pre-filtering a soundfield with the near-field compensation filter in anticipation of decoding to loudspeakers located at a pre-determined radius. This radius is the reference radius. We can view this as the anticipated loudspeakers finishing off the synthesis of the curve of the encoded waves by physically adding the remaining curve.

This is all good, but the true genius is the inclusion of a the near-field effect control filter, which combines the near-field compensation and proximity filter into a single block. The arrangement is as illustrated above, but without the inclusion of focus. Also, instead of having a single distance argument, there are two, one for each filter. Doing so makes it very easy to translate the reference radius. In other words, we can move where basic waves are encoded, which easily facilitates radial beamforming."

What seems to be missing for me in this explanation is the prerequisite of knowing what the problem actually is that HOA-NFE solves here. Why is it nice that Daniel translated the basic wave to a radius other than inf? In other words, what's the selling point for NFE in (something slightly closer to) layman's terms? I think this might clear up the explanation, at least for me.

Thanks again!

best

[madskjeldgaard]

The abstract from the Jérôme Daniel paper might be helpful in clearing this up:

Higher Order Ambisonics have been increasingly investigated in the past years, and found promising as a rational, scalable and flexible way to encode, transmit and render 3D sound fields. Nevertheless, studies concerning virtual source imaging or natural 3D sound encoding mainly focussed on the directional encoding of plane waves, and neglected the near field effect of finite distance sources though it's present in any ordinary sound field. This paper highlights that with near field, the infinite bass-boost affecting ambisonic components makes the currently accepted format unviable. By introducing from the encoding stage a near field compensation of reproduction loudspeakers, a viable, modified ambisonic format is defined, distance-coding filters are designed, and higher order ambisonic recording and synthesis become practicable.

[joslloand]

Hello @madskjeldgaard,

Yep, the problem is, if we directly use infinity referenced proximity, we get super high gains when sources are in the near-field. Daniel's NFE solution is to pre-filter to a radius in the mid-field. We can still blow things up with near-field sources, but we have a lot more room (radius, actually!) to work with.

Good idea to include some more info regarding Daniel's comments.