True monitoring: room for reflection

In this week's tech note Professor Rumsey reveals why listening rooms very rarely sound the same and explores the effectiveness of room correction systems.
Author:
Publish date:
29842.jpg

The rooms in which we reproduce audio signals for sound mixing are not anechoic but have surfaces that create reflections. This is what gives a room its unique acoustic ‘signature’ and it is rare to find one room that sounds exactly the same as another. While there may be claims about ‘neutral’ rooms, in reality there are just different forms of acoustic modification of what comes out of the loudspeakers. Some of them are more benign or desirable than others. Having been in numerous ‘standard’ listening rooms, I can confidently say that they all sound different despite nominally conforming to an international standard. I realise this is a controversial subject but there seems to be a growing realisation that natural stereo reproduction is not about suppressing all the reflections but about exciting them in a more uniform way that enables the brain to distinguish between what’s coming out of the loudspeaker and what is being reflected from the room. If the off-axis loudspeaker response is way different from the on-axis response then any room reflections will be very different from the direct sound and the brain may find it more difficult to ignore them. If you’ve ever been in an anechoic chamber you’ll know how weird the experience is. It’s not something we encounter in natural listening very often and there’s a strange contradiction between the visual experience of being in an enclosed space and the auditory experience of being outside, suspended a long way above the ground. One thing’s for sure – even though anechoic chambers do away with the thorny problem of room modes and reflections, you wouldn’t want to mix an album in one. The people that ultimately consume your mixes don’t listen in anechoic rooms either, so there’s no point in you trying to avoid reflections in your mixing room. In acoustics classes you get taught that there are room modes, which are pressure peaks and troughs in different places in the room, arising from standing waves that get set up between reflecting surfaces. At low frequencies they are usually quite widely spaced, so they tend to stand out more noticeably, especially in small rooms. This can give rise to ‘boominess’ when a particular frequency rings on long after others have died away, and there are various acoustic solutions to such problems such as resonant absorbers and bass traps. There’s been quite a bit of research into so-called ‘ideal’ room ratios, which are relationships between room dimensions intended to ensure that modes are distributed fairly evenly in space and frequency. Cubic rooms, or those with dimensions of simple multiples, are the worst because the modes in each dimension tend to coincide. It’s a bit of a myth that making your walls non-parallel gets rid of the problem – it just makes the mode frequencies more difficult to predict, although it might avoid the worst axial modes. Recent research by John Sarris found that Trevor Cox’s general ratio of 1:2.19:3 performed reasonably well when evaluated against a variety of metrics. The EBU’s guideline of 1:1.96:2.59 also worked well, and its limits for ‘good’ ratios avoided the worst problems. Room correction systems such as Trinnov’s Optimiser use signal processing in an attempt to compensate the loudspeaker/room response. This can reduce the spectral variations caused by room modes, as well as introducing a deconvolution process that aims to cancel the effect of early reflections to some extent. The degree to which this should be done is a matter for experimentation, and depends a lot on the problems in your room and the flatness of your monitor’s off-axis response. It may help, though, to make mixes more ‘translatable’ between mixing rooms if they all employ similar compensation systems. 

Related