Changes in weather, wind and humidity may have a great impact on our equipment.

Many sound engineers are very busy outside, running around in various performances, exhibitions and festivals. We have to clean the filter of the amplifier every day, put the return speaker in a trash bag to keep it dry, and cover the mixer with a curtain (also to keep it dry).

In other words, the biggest challenge we face in outdoor performances is the impact of atmospheric conditions on system performance. Changes in temperature, wind, and humidity can severely damage our carefully orientated and debugged systems.
The larger the site, the greater the impact of atmospheric conditions on sound transmission. These effects cannot be prevented, but at least they can be predicted (or partially predicted).

Morning, noon & night

Whenever you are doing audio work outdoors, temperature gradients are a problem. In the morning, the ground maintains night temperature for longer than the surrounding air, resulting in a layer of cold air near the ground and warmer air above.
As the temperature increases, the speed of sound increases slightly. For example, when the altitude is zero feet and the temperature is 50 degrees Fahrenheit, the sound propagation speed is 110.7 feet/100 milliseconds.
If the temperature rises to 90 degrees Fahrenheit, the speed of sound will travel 115.14 feet/100 milliseconds. The sound propagation path of the loudspeaker will be slightly downward, bending towards the cold air layer.

In more extreme cases, the sound wave may bounce on the ground, jump over part of the audience, and then refract downward again, causing a blind spot in the coverage of the system.
The situation is just the opposite at night. The surrounding air cooled down, but the ground was still warm, and a layer of hot air stayed near the ground. Therefore, the sound propagation path will be upward and may be refracted directly above the crowd. (Please note that the warm air generated by the crowd exacerbates this trend.)
Due to the relative humidity in the air, the energy loss as the distance increases
Wind has a similar effect. The speed of sound propagating in the wind is equal to the sum of sound speed and wind speed, so when sound travels on the wind, the wind speed must be subtracted.
Since the speed of the wind in boundary areas such as the ground is zero or almost zero, the wave front facing the wind will be refracted upward, because the top of the wave front is affected by the head wind, and the speed will be slightly slower.
If the wind pushes the sound forward, the sound wave will bend downward. It is not the wind itself that causes these problems, but the wave speed that varies with altitude. The influence of crosswind can be analyzed by simple trigonometry. (Is it true?)
Let's look at an example. Everyone knows that the nominal speed of sound is 770 miles per hour. Suppose the crosswind is blowing at a 90-degree angle in the direction of sound propagation in the sound system, and the speed is 40 miles per hour.
We can think of these speeds as the two sides of a right-angled triangle to get the angle of refraction. In this example, the refraction angle is approximately 6 degrees.
However, this method can easily be misleading. Because the cover angle of the speaker group is usually 120 degrees or more, the moving direction of some wave fronts is perpendicular to the wind, but the wave fronts of other parts may only be in the same direction as the wind, or far away from the wind at all.
So, although the wind is pushing the sound forward, their behavior is affected. All in all very complicated!

Influence of humidity
Humidity is another factor that greatly changes the sound transmission of the audio system, but its influence is mainly reflected in the frequency domain.
Although it sounds inconsistent with our intuition, the lower the humidity, the greater the sound attenuation; the higher the humidity, the less the sound attenuation.
The influence of humidity on the frequency response starts at 2 kHz, the higher the frequency, the more obvious the influence becomes.
If the distance is 100 feet and the humidity is 20%, 2 kHz will attenuate 1 dB, and 10 kHz will attenuate as much as 8.5 dB.
And as the distance increases, these energy attenuations will accumulate. If the distance is 200 feet, the 10 kHz attenuation will double to 17 dB! Moreover, these energy attenuations are not included in the attenuation of the inverse square law. They are not linear with frequency, so the amplitude response of the coverage area may vary greatly.
The inconsistency caused by 10%-40% humidity is the most obvious. After that, as the humidity increases, the energy attenuation will become smaller and become more linear throughout the frequency range.
If the array is composed of point sound sources and the total coverage of the vertical surface reaches 50-80 degrees, then the above factors may have a relatively large impact. However, if it is a linear array, the wavefront on the vertical axis becomes very narrow due to the interaction, and the probability of error in directivity is relatively low.
In line arrays, compared to low frequencies, high frequencies can maintain the boasted law of "double the distance and 3dB energy attenuation" over a longer distance. But this phenomenon has been compensated, because high frequencies are more susceptible to the effects of climate and energy attenuation. However, the energy attenuation caused by humidity is not linear and may not be as useful as expected.
Line arrays are usually used to cover larger venues, and the phenomenon we are discussing here will become more obvious as the distance increases.
The more air the sound wave needs to pass through, the more likely it is that negative effects will occur. At a distance of 100 feet, the impact becomes obvious. If the distance reaches 500 feet, the impact will become very significant.

The main solution
How can we overcome the adverse effects of climate factors? One way is to use delayed stacks. But you might say that those are technologies of the 20th century. Can't line arrays eliminate them? This is not necessarily true.
In the war against temperature and humidity, bringing people closer to speakers is a key weapon. This will not only achieve the desired frequency response, but also maintain a uniform volume distribution in a relatively large area.
Indeed, the use of a delay system faces very troublesome mechanical adjustments.
Obstructing the line of sight, audio feed and power supply issues, and more time for construction and disassembly increased the cost of the performance and made the performance process more complicated.
However, we can minimize these inconveniences. Since air absorption has less impact on low frequencies than high frequencies, we can skip the subwoofer. In some cases, you can even skip the low-frequency cabinet in the delay system.
This greatly reduces the size of the system and reduces power requirements. Co-locating the delay source and the mixing position can alleviate audio and power feed issues.
Some new small line array systems are very suitable for use as delay systems. They provide enough output in a relatively compact volume without blocking the line of sight. Alternatively, a full-range speaker with a smaller size can also be used. What is the most appropriate distance between the delay speaker stack and the main amplifier speaker group? This sometimes depends on the consideration of the physical structure, and sometimes it depends on the sound pressure level limit of the venue (taking into account the surrounding community and other influencing factors).
If the sound pressure level is measured at the main amplification pitch, the main amplification system can operate at a lower level, and the delay system does not have to be far away from the stage. Modeling programs or simple mathematical operations and the inverse square law (or only the inverse square law in the case of online arrays) can be used to determine the acceptable level attenuation before the signal needs to be amplified.
Remember that the additional attenuation described above is not included in the theoretical attenuation. If the performance is held in an area with clear weather and high humidity, the energy attenuation caused by the environment may not be obvious. But if the show is held in a windy desert, be careful!

Appropriate signal delay
How to find the most suitable signal delay? I think measuring the actual time difference is the best way.
The emergence of line arrays makes delay speaker stacks dispensable? But in fact, it's not
Use Smaart or TEF to generate impulse response or energy time curve (ETC). This can clearly display the arrival time of the main amplifier system and the delayed speaker stack, allowing you to obtain a delay number through the cursor.
If you don't have these tools, you can get the delay time through mathematical operations. At 70 degrees Fahrenheit and zero altitude, the speed of sound is 1130 feet per second, or 0.88 milliseconds per foot. If you know the distance, you can calculate the delay time.
Many sound engineers like to use the Haas (priority) effect. The human ear locates the sound based on the arrival time and frequency content. The sound that arrives first and/or the sound that carries the highest frequency content will determine the human ear's perception of the sound direction.
The human ear can also synthesize sounds that arrive within a 20-millisecond time window, which is called the Haas region. In other words, within this time frame, the ear will not perceive signals with different arrival times.
As a result, the signal can be slightly delayed from the correct acoustic settings and the high frequencies can be cut slightly to convince the audience that all sounds are coming from the stage system. This is called sound localization. If the audience reports that the delay speakers are not working, and you know that it is not, it means that everything is set up correctly.
And don’t forget, the speed of sound changes with temperature. If there is a big temperature difference in the area where the performance is located, please reset the delay so that it is as close as possible to the performance time.