Originally Posted by Martycool007
What is the correlation from frequency problem(s) to using either diffusion or absorbion? Do you mean like peaks and nulls in certain frequencys? I have never used REW so I am not well versed on what possible problems might occur, but I am in the process of obtaining all of the necessary equipment to get started using REW, so I am trying to learn some of this stuff now. Are there other factors that come in to play, other than peaks and nulls? I have read that the whole RT60/reverb thing is unimportant. This came from a guy whom is an expert in room acoustics.
Sorry for getting off topic!
Is it better to treat with diffusion when dealing with mid to high frequency problems? Ie, 400hz and above?
yes, rt60 (reverberation time) is irrelevant measurement in Small Acoustical Spaces (SAS) as there is no statistically developed reverberant sound-field at any frequency we are concerned with. what little reverberation exists is above our hearing range and below the ambient noise floor. we do have LF decay and specular room decay, but many incorrectly (sloppily) mis-apply the term 'reverb' to such behavior. measuring rt60 also implies the requirement of an omni-directional source (dodec), of which typical home loudspeakers are NOT satisfactory. it also is to be measured past Dc (critical-distance), of which we do not have in a Small Acoustical Space (See below).
in Large Acoustical Spaces (LAS), there is a developed reverberant sound-field that is homogenous (equal and probable in all directions) - this is frequency dependent based on volume of the acoustical space (thanks to Dr. Manfred Schroeder). there exists also in LAS a 'critical distance' of which the reverberant sound-field becomes louder in gain than the direct signal. in a LAS, the reverberant sound-field becomes the effective noise-floor. also note in LAS, echos are perceived (ignoring gain for simplicity) as a reflection that arrives > 80ms (that's ~89ft of travel!). how large is your home listening room?
this is not the same acoustical behavior we experience in Small Acoustical Spaces - as in SAS we are subject to focused specular energy (reflections) that bounce around the room of which the vector (ingress direction of the signal), magnitude (gain), and time-arrival can all be resolved. the energy is NOT well-mixed and response changes drastically as you move throughout the room. the frequency response at a given point in 3space is determined by the summation (superposition: http://en.wikipedia.org/wiki/Superposition_principle
) of the direct and indirect signals. since the indirect signals arrive at later times than the direct signal (due to longer paths of travel and with the speed of sound being a constant within your room's medium), they will sum constructively and destructively with the direct signal producing a comb-filter interference pattern within the frequency response (peaks and nulls; acoustical interference). a comb-filter doesn't exist in the real world but is simply an interference pattern manifested within the frequency-response (important distinction to note).
it is also important to note that REW does not tell you whether you need absorption, diffusion, redirection, etc or anything of that nature. measurement tools do not make these decisions for you but help you understand the actual (measured) acoustical behavior of the space based on source-receiver position and how that changes as you apply treatment in an attempt to achieve the desired response.
understand that there is a modal region of which the acoustical energy has wave properties - where the wavelengths are larger than the room boundaries. a transition region (as a wavelength may be larger than one boundary/axis but not necessarily the other 2) - and then the specular region where the acoustic energy behaves and can be modeled like ray-tracing light (angle of incidence = angle of reflection; geometry exactly like how pool balls bounce around a billiards/pool table). for the modal (LF) region (typically 0-250hz), you can utilize the Waterfall Plot within Room EQ Wizard. this will detail to you the frequency response of the LF region as well as the decay times (displayed as the Z-axis). this will indicate to you how long it takes acoustical energy at those frequencies to decay down. you will usually see high decay times from frequencies relevant to room modes, as those are resonances within the acoustical space as the energy tends to persist. long LF decay times will impart a perceived response at the listening position of the bass notes running together into a muddied mess.
since the specular region (~250-300hz and above) can be modeled like light, this is why you can utilize a mirror to identify 'reflection points' - but the mirror does NOT detail any relevant information to the acoustical energy incident from such a boundary. within Room EQ Wizard, there exists a tool called the Envelope Time Curve (ETC) response, which displays the time-domain behavior and how specular energy impedes the listening position. the frequency response details you Gain vs Frequency, while the ETC displays Gain vs Time, and time can be substituted with Distance since the speed of sound will be a constant. you can think of this like an impulse such as a gun-shot or balloon bursting. the direct signal will arrive first in time due to the direct signal to listening position being a straight vector and thus the shortest path traveled. following the direct signal will be any indirect energy (eg, first-order early reflection paths), followed by the later arriving reflections, and then the remaining energy as it decays until the last of the energy has been fully damped. the longer the reflection path, the longer in time it will take for that energy to reach the listening position (mic).
the ETC allows one to identify ALL of the destructive specular interference, and work backwards to resolve the particular boundary that the energy is incident off of. for example, if you measure a high-gain reflection that arrives 6ms after the direct signal, you know that the reflection path traveled (6ms * 1.126ft/ms) ~ 6.756ft longer than the direct signal. you can then work backwards and identify exactly which boundary this measured high-gain reflection is incident from and apply treatment as necessary. you're essentially identifying the ACTUAL high-gain early arriving reflection paths vs merely all POSSIBLE reflection paths as with a mirror. thus, you're not blindly placing treatment (absorber, diffuser, reflector) but instead placing the treatment at actual problem areas. eg, why would you place a large, expensive diffuser at a reflection point that is not incident of any high-gain reflections? and to the same point, many place absorption panels blindly via the mirror at any and all reflection points, which can quickly lead to a dead room. instead, if your criteria is to attenuate any early arriving high-gain indirect signals then you would place broadband absorption surgically ONLY at the areas incident of high-gain reflections (as measured with the ETC) and limit the amount of broadband absorption within the room.
but the ETC is merely a measuring tool and does not indicate the particular response you wish to achieve. that is your decision. and diffusion is for the specular region and has its own set of design criteria based on a multitude of factors and variables. for a 2ch listening space, you would want 1-Dimensional Reflection Phase Grating diffusers (QRD/PRD) with the wells oriented vertically such that the diffused returns off the rear wall are dispersed in the horizontal plane - which allows the diffused returns to arrive laterally from the rear side-walls for envelopment. bandwidth would be extend as low in frequency as physically possible, as the diffuser needs to be large with respect to wavelength as well as minimum seating distance requirements for Reflection Phase Grating diffusers (near-field vs far-field). so seating distance to rear wall is a factor in design as well.
edit: and reflection phase grating diffusers are absorbers for all intents and purposes -