Hi all! I hope you are doing fine. I’m finally back with the third article in this series.
After discussing the prerequisites and the acoustic design concept in the first two articles of this series, we will now have a look at the first measurements of the naked room that finally became our new control room. As you can see from the following pictures, the room is a typical cellar room.
It’s immediately obvious from the pictures that symmetry, one crucial part of room acoustics, might be a problem in this room. On the one hand, there is the chimney and the door in the rear left corner as well as the ventilation pipe at the left wall. On the other hand, there is a naked rear right corner and right side wall. We will see what these things do to the frequency responses of left and right speakers respectively.
The first crucial step is to determine the best starting position for the speakers. This is done by performing several measurements starting at a reasonable position and successively varying the position of speakers and microphone for each measurement. Finally, the measurement that gives the flattest frequency response is used as the best starting position. In small rooms it has been shown several times, i.e., often measured in the recording.de room acoustics forums that the speakers are best positioned directly at the front wall, i.e., with the smallest possible gap to the front wall. This is due to the speaker boundary interference response (effect) (SBIR/E) producing dips in the bass frequency spectrum due to destructive interference of the direct sound from the speakers with it’s reflections from the front wall (the wall behind the speakers). If the speakers are positioned close to the front wall the dip in the spectrum is moved to lower frequencies which in turn can work against room modes, resulting in a better overall frequency response. Be aware though that this is only beneficial in small rooms where it also helps to maximize the distance between listening position and rear wall. By the way, the SBIR effect is also the major reason why main monitors are often flush mounted in pro studios.
Nevertheless, the best position always has to be determined by measurements since no two rooms will behave the same and therefore an acurate prediction of the best position is quite difficult. You can see pictures of the best position as well as it’s frequency response in our room in the following pictures.
As you can see, there are massive peaks and dips in the frequency response. These are due to room reflections superimposed on the direct sound and thus altering the perceived frequency response via comb filtering. In the course of treating the room to achieve our target values (RT60 etc.) we will see that the frequency response will automatically start to flatten due to the elimination of room modes and reflections. At the moment we have a difference of over 40db between the highest peak and the lowest dip, which is quite horrible.
The early decay time (EDT) curve gives a first impression of how disastrous the reverberation of the room is. This is the curve that corresponds to the RT60 value, we calculated last time using Bob Gold’s calculator. (We won’t call it RT60 here, because that terminology is only valid if we had a completely diffuse sound field.) Remember, that our target value is between 200ms and 400ms throughout the whole audible spectrum. At the moment the curve jumps between approx. 3600ms and 900ms, with the worst area between 100Hz and 200Hz. This resembles the modal problems that Bob Gold’s calculator computed, with the most massive room modes up to 200Hz.
These problems are even more obvious in the so-called waterfall diagram. In that 3D graph the frequency is plotted logarithmically on the horizontal axis, the level in dB on the vertical axis and decay time in ms on the axis perpendicular to those two. If you trim it to show a level difference of -60dB from the highest peak, then you have another depiction of the early decay time. In a proper control room all frequencies should have been attenuated by 60dB within 200-400ms. If we take a mid-range value of 300ms this means nothing should be cut off in the first waterfall plot. This obviously isn’t the case with our naked room.
To further elucidate the problem I show you a waterfall plot with a decay time setting of 2000ms (2s). Even here we still have cut off parts in the modal region between 100-200Hz. So there clearly is a massive problem in this region.
The spectral plot shown in the next picture is another alternate view on the rooms decay time, showing the frequency on the horizontal and the decay time on the vertical axis as well as the level coded as different colors.
The last important plot is the Energy Time Curve (ETC) providing information on the early reflections and their temporal location, i.e. the time difference between direct sound and a given reflection. This will allow us to find the points of reflection, e.g., side-walls, ceiling, etc.
An explanation of the ETC of an untreated room like ours and a perfect example for a proper ETC can be found in the next picture. Our target is to achieve an ETC that is as close to the latter picture as possible.
As you see, an untreated room has massive problems that have to be cured in order to use it as a proper control room. Otherwise, you would never be able to make objective decisions for your recordings or mixes, resulting in mixes that don’t translate to other systems and places.
I hope you will be part of this journey of transforming the horrible room we just discussed into a proper control room where making objective decisions is no longer a whishful dream but rather an actual fact.
More on that next week, when we will start to build bass traps and measure their impact on the rooms response.
Best regards and keep the music coming,