声学中长期存在的问题之一是预测房间内的声压级。解决这个问题的传统方法依赖于混响场的概念，假设混响场在整个房间内是均匀的。然而，多年来大家都知道，在许多情况下，这一假设会被违背，并且相应地，声压级的预测也很困难。例如，在其他高反射环境中，高吸收性材料的贴片可能会导致预测方法出现问题。 造成困难的另一种情况是“不相称”的房间，即尺寸比超出1:1-5:2限制的房间，这是产生合理漫射场所必需的。这样的房间有很多例子。大型开放式办公室是一个常见的例子，许多工厂也是如此。本文将关注在这样一个不相称的房间内预测声压级。它将特别关注大型开放式办公室天花板空调通风口引起的声压级预测。 对于大型开放式办公室的声压级预测问题，将讨论三种方法。这三项都是基于计算机计算的。前两种方法需要大型计算机设备，本文主要用于比较。第三种方法与扩散场假设不同，将房间视为一对吸声平行平面。这种方法更简单，更容易编程，并且给出的结果与大型低天花板房间的测量结果一致。这三种预测方法均源自几何声学推理。 “射线”或几何声学方法和波浪方法各有其优点和局限性。虽然可以从波动法中学到很多东西，但本文将关注几何声学法以及基于它的方法。引文:俄亥俄州辛辛那提ASHRAE学报第87卷第2部分研讨会

One of the persistent problems in acoustics has been the prediction of sound pressure levels within rooms. The conventional approach to solving this problem has relied on the concept of a reverberant field, assumed to be uniform throughout the room. However, it has been well known for years that there are many conditions under which this assumption is violated, and where the predictions of sound pressure levels are correspondingly difficult. Patches of highly absorbent material in an otherwise highly reflective environment, for example, can cause problems with prediction methods. Another situation which causes difficulty is rooms which are "disproportionate" that is, rooms whose dimension ratios fall outside of the 1:1 - 5:2 limits which are necessary to produce a reasonably diffuse field. There are many examples of such rooms. Large, open-plan offices are a common example, as are many factory spaces. This paper will be concerned with the prediction of sound pressure level within such a disproportionate room. It will be particularly concerned with the prediction of sound pressure levels caused by ceiling air conditioner vents in large open-plan offices.Three approaches will be discussed to the problem of predicting sound pressure levels in large open-plan offices. All three are based on computer calculations. The first two require large computer facilities, and are presented here primarily for comparison purposes. The third method departs from the diffuse field assumption and treats the room as a pair of acoustically absorbing parallel planes. This method is simpler, easier to program, and gives results which appear consistent with measurements in large, low-ceilinged rooms.Each o£ the three prediction methods is derived from geometric acoustics reasoning. The "ray," or geometric acoustics, approach, and the wave approach each have their strengths and limitations. While there is much to be learned from the wave approach, this paper will be concerned with the geometric acoustics approach and the methods based on it.