(sorry for the click bait title – wish everything was that simple, but I do encourage you to read)
A Continuation of Original Topic at:
Reverberation and Clear Speech Standards
We may all know that the effect when sound lingers after the source has stopped generating sound is called reverb. Reverb happens when sound reflects of other surfaces and reaches our ears at different times. We have already discussed the resonant effect of reflections (room modes), reverb is yet another side effect of reflections.
Human ear is able to intelligently interpret Phonemes (basic unit of intelligent speech) within 25 to 60 milliseconds. Meaning speech sounds must be at least 25 milliseconds apart for the ear to decipher the sounds separately. Anything before that is seen as uni-source and unidirectional (Coming from a single source from a single direction). Before the 25 millisecond barrier, the precedence effect or Haas Effect is observed - as in if 2 similar sounds arrive within 25 milliseconds of each other, the ear will not be able to tell the sounds apart and the direction of the sound will appear to come from the sound that arrived first, even if the sound that arrived later was relatively much louder than the previous sound. In most cases, this effect is considered a bad thing for speech intelligibility, however this phenomenon can be exploited to create a richer depth in sound mixes where a little intelligibility can be sacrificed for the sake of musical and acoustic nuance (but that is a topic for another discussion).
Coming back to the basics of reverberation and clear speech -
Contrary to phonemes, syllables can take upto 200 milliseconds for the ear and brain to decipher. In theory a clear speech broadcast must be at least 25 milliseconds apart and must not linger past 200 milliseconds. (Hence the Telecommunication Union and European Broadcast standard for 200 ms RT60 reverb time for broadcasting control rooms)
The International Telecommunication Union (ITU) standards are often extended to music rooms and mixing rooms. As in most mixing control rooms target to have an RT60 Reverb time of under 200 milliseconds.
RT60 Reverb time is called the time for a sound signal to decay by 60db in a large room after the sound source stops. While RT60 is mostly relevant in larger rooms, RT60 standards are still often extended to smaller rooms. In smaller rooms early reflections within 10 to 15db matter more than the overall reverberant field of sound of 60db. Which is why smaller rooms focus more on treatment of early reflections and larger rooms focus more on critical distances and the overall reverberant field.
Lets talk about RT60 a bit more
Wallace Clement Sabine, a harvard based acoustic engineer of the boston symphony hall established a relationship between the reverb time RT60, the volume of the room and total surface absorption of the room.
Today we deal with absorption in units of Sabin.
Absorption of a surface in Sabins equals the area times the absorption coefficient of the surface.
Each building material has a sound absorption coefficient measured from 0 to 1 Sabins per sq ft in most cases ignoring sound diffraction from edges (another topic of discussion).
For example a solid wooden door has an absorption coefficient of 0.14 at a frequency of 125 Hz.
The formula empirically derived by Sabine is
RT60 = 0.049 x Volume of Room in cubic feet / total absorption of the room in Sabins
Total absorption = sum of Area of each surface multiplied by absorption coefficients.
Octave Bands and Absorption:
In acoustics, we always deal with sounds in “Octave Bands”
31.5, 63hz, 125 hz, 250hz, 500hz, 1kz, 2khz, 4khz, 8khz, 16khz
Each frequency has a different reverberation time, this is due to the fact that each frequency has different wavelengths and different travel times as discussed previously. Ideally (according to broadcast standards) the goal is to achieve similar reverberation times in each frequency band and in a perfect world, around 200ms or 0.2 seconds.
Lets take an example of a brick and mortar room with a concrete floor with dimensions:
14 ft by 14 feet by 10 feet.
The room has a volume of 1960 cubic feet, hence the total absorption needed for room to have a Target RT60 of 0.2 seconds is 412 Sabins.
Total Absorption needed to achieve the broadcast standard = ( 0.049 * 1960 ) / 0.2 = 480 Sabin
Now Lets calculate how much actual absorption is in the that room:
The absorption coefficient of Brick and Mortar is around 0.1 around 500hz
The absorption coefficient of concrete floor is around 0.05 around 500hz*
*material-data-CERN.pdf (959.0 KB)
Total Absorption of Concrete floor = 14x14 * 0.05 = 9.8 Sabin
Total Absorption of The Brick and Mortar Ceiling = 14x14 * 0.1 = 19.6 Sabin
Total Absorption of 4 Brick and Mortar Walls = 4x (14x10) * 0.1 = 56 Sabin
Total Absorption of the room = 9.8 + 19.6 + 56 = 85.4 Sabin
As you can see that we are about 400 Sabin Absorption short of achieving our goal of 0.2 second Reverb time which means we need to find absorptive material to cover for about 400 Sabin of Absorption. This is in addition to dealing with room modes. So ideally you want to find absorptive material closer to your problem room modes while still covering about 400 Sabin worth of absorption in the above mentioned room.
Reverb in small rooms
In smaller rooms we deal with early reflections (within 15 milliseconds to 25 milliseconds) to minimize comb filtering we discussed earlier. Please be advised that this is not the same as flutter echo as commonly misunderstood. A flutter echo is separated usually by more than 60 milliseconds, it is a strong delayed reflection between 2 parallel walls. Flutter echoes tend to go away as we treat smaller rooms for early reflections but the treatment itself does not focus on flutter echoes, it focuses on comb filtering.
Broadcasting Union suggests that a signal should decay by 10db (in some cases 15db) within the first 15 milliseconds for certain frequency ranges.
The mirror trick:
To find early reflection zones, put a mirror on each side of the wall and ceiling and if you see your speakers in the mirrors from your listening position, that will be your point of first reflection and the first to be treated. Absorptive or scattering material is usually required at this location. With absorption we target for sound level to drop by 10 db or more and for scattering the goal is to make the sound travel 17 feet or more before reaching our ears.
Why 17 feet? its simple, divide 17 feet by speed of sound 1130 and you get 15 milliseconds (the broadcast recommendation)
Those are some guidelines at a very high level. Of Course, it depends on the purpose of the room. If you want a room to sound reverby, you can control how much absorption you add so you can get your target reverb times. This is only recommended for tracking instruments and not necessarily for control room mixing. You dont want natural reverb times of the room messing around with the reverbs you added with the plugins tarnishing your sound image. For control rooms, the reverb times should be rather tamed to below 200 milliseconds as recommended by the broadcasting standards.
That’s all for now folks, next up is Sound Diffusion.