"Improving Assessment of Low Frequency Room Acoustics"

[twelti]

This is going to be a multipart review of a very interesting Audio Engineering Society (AES) paper that attempts to correlate some basic room acoustic parameters to subjective impressions for low frequencies in small rooms. For those inclined to read the full paper, it is:

"Improving the Assessment of Low Frequency Room Acoustics Using Descriptive Analysis", M. Wankling, Bruno Fazenda, and William Davies. AES Convention Paper 8311, November 2010. These same authors have done some other interesting and competant work in this area as well.

You can purchase this and related papers for a nominal fee from the AES website:

http://www.aes.org/e-lib/online/search.cfm

My aim here is to make the core results more accessable to interested readers who may not be familiar with the more acedemic methods, such as "Decriptive Analysis". I also offer my own commentary in addition to summarizing the authors' work.


In short, what they did was twofold:

• Using a group of trained listeners, agree on a small but robust set of subjective verbal descriptors that together determine overall preference.
  
• These descriptors were used in listening tests, to investigate the relationship between real acoustical room parameters, the subjective descriptors, AND overall preference.

,

So, if you are interested in the relationship of, for example "articulation" with room volume, and with overall subjective preference, this paper covers exactly that. Having read many of these sorts of acedemic papers, I can say that this one is very well done.

Next post will cover how the subjective verbal descriptors were determined, and what they are. Keep yer suspenders on...!

[Steve Williams]

Looking forward to reading it Todd

[twelti]

The first part of this very interesting research involved generating a number of verbal descriptors for subjective quality of bass. They use a corps of trained listeners, and a fairly well known method called "descriptive analysis", which incidentally began in the food industry. The process is all quite acedemic, but in short you have a group of listeners and you try to come to a consensus about a minimum set of descriptors which fully describes all subjective aspects of "bass". One could debate the validity of this process at length, but in fact the descriptors they came up with sound very reasonable. The descriptors, with corresponding definitions are:

ARTICULATION
"Muddy/Tight"

"Muddy": sound (or note) has a lack of definition, and could sometimes be descried as "smeared".
"Tight":Each sound (or note) is destinct, well defined and precise

RESONANCE
"None"/"High"

A resonant sample has or doesn't have notes which sound louder, ring and last longer

STRENGTH
"Weak"/"Strong"

Relates to the loudness fo the low frequency when compared to the rest of the frequency range in the sample

DEPTH
"Shallow"/"Deep"

Has or does not have notes that extend down lower in frequency.

SO, I submit that these are pretty reasonable terms, though could be debated. Anyone wanna...?

Next installment, I will discuss the actual listening tests they did using these descriptors.

[Nyal Mellor]

Could also use EXTENSION instead of DEPTH...? Otherwise I like them