PERCEPTIVE / ORGANIC EQS

 

1. Principle 2. The micro aspect 3. Changes over time 4. Changes according to sound level 5. Would be similar to... 6. How to make such an EQ 7. Summary 8. Example

 

1. PRINCIPLE   This is an EQ as we commonly know it :   1. What people usually consider as important in EQs is exactly what we see on the picture : the ability of removing or boosting certain frequencies. The "micro" aspect is completely overlooked. Looking back at the chapter "perceptive / organic processing", the micro aspect is, on the contrary, something important.   2. This EQ uses algorithms that are designed to remove / add gain, more or less broadly around certain frequencies. Howener natural this looks, it's more a visual / intellectual process than a perceptive process : the mind finds an easily understandable concept, and uses mathematical algorithms to make this concept true - to transpose this concept into reality. It remains to be seen whether such a process - adding gain around certain frequencies using common algorithms - corresponds or not to anything that makes sense in reality. ("Making senses in reality" : a straight line doesnt make sense in reality, it's a purely intellectual object, though it can find practical applications. It doesnt exist. A tree, a coastline, are not really understandable by the mind but they exist in reality - see "perceptive richness" for more ideas on this matter.

2. THE MICRO ASPECT   This is a particular kind of EQ : the EQ the external ear and head apply to the sound, depending on the sound source position : (it's often referred to as HRTFs, Head Related Transfer Functions)   Given a fixed sound source azimuth, the curve looks like that : The global shape is not easy to grab, neither is the "micro" shape. As in natural shapes (typical example, a coastline) it is obvious that zooming in would only bring more details. NB : it is completely uninteresting to know if the curve has a fractal structure with shape similiarities at different scales - the interesting point is just that it has a lot of details at different scales

3. CHANGES OVER TIME   When the sound source position changes, so does the EQ ... but as the head does micro movements all the time, the EQ changes all the time, though slightly => the EQ detail changes over time NB : in real life, those "micro changes" help source localization - but this particular point is not of much use in the present case

4. CHANGES ACCORDING TO SOUND LEVEL   If we look at the Fletcher curves, which include the influence of HRTFs amongst other things, we can see that the ear EQ changes drastically depending on the sound level. This proves to be enormously useful when it comes to communicate (see audio examples). Apart from this fact, it's a property that is assimilated by the brain to sound "natural". It should, or at least can, be part of the concept of a "natural sounding" EQ.

5. WOULD BE SIMILAR TO... Another display of similar EQs :

6. HOW TO MAKE SUCH AN EQ   The HRTF, the EQ that comes from the external ear, is a complex delay unit (direct sound + numerous reflections all over the head and in the ear conduit). From a spectral point of view, HRTFs are a complex sum of comb filters. The first step to make EQs that emulate HRTFs would be to build a very complex, short timed delay unit. Delay parameters should always vary over time, though slightly. Is there a point in trying to emulate the "macro" aspect of the Fletcher curves ? Maybe, maybe not : Cons : evidently, the Fletcher response is calibrated on the ear's dymanic range - doubling them would produce too much level difference between frequencies. Pros : if it's the brain that needs this Fletcher EQing rather than the cochlea, then even more emphasizing the frequencies than the brain needs can be a good thing.

7. SUMMARY   What do we need here to have an organic EQ : 1. very complex delays, very short durations (measured in micro seconds probably) 2. delay parameters keep on moving slightly all the time 3. from a macro point of view. global shape changes according to level (typical : 3.5k boost at low levels) It is worth noticing that in the audio field, such an EQ would not be considered as acceptable : thinking about something as "un-pristine" as a complex, short timed delay unit to be used as an EQ would immediately be suspected to "ruin the signal" etc. ... but it is the same kind of EQ that enables the ear to be so astoundingly precise and accurate.

8. EXAMPLE   This is how a boost around 3kHz would look :   for a given level for a higher level .... and with the irregularities moving all the time (maybe randomly   Back to main index