Practical information, tips and warnings

•

When measuring room impulse responses using WinMLS, it is recommended that a room acoustics setup is loaded. Such a setup can be loaded and saved from the Setups Toolbar (can be turned on/off from the View menu), or from Measurement->Save/Load Measurement Setup. When the setup is loaded, the GO button on the standard toolbar can be used to start a measurement, or File->Insert can be chosen to insert an already existing response from a file.

•

Calculation options and parameter settings for the room acoustics calculations can be modified from the Room Acoustics menu. These settings are saved with the post-processing setup file found on the Setups Toolbar.

•

The room acoustics calculations have been developed with accuracy as the main goal. Optimization of calculation speed has not been allowed to compromise accuracy. The calculations have been tested against international benchmarks with excellent results. (Post-processing tests have been carried out on impulse responses described in a report by J.S. Bradley, “An international comparison of room acoustics measurement systems,” National Research Council of Canada (1996)).

•

The room acoustical parameter results are given with a considerably higher accuracy than the typical measurement repeatability when repeating measurements in a given position in a normal room.

•

To avoid time consuming calculations, do not permit excessive impulse response lengths and sampling rates.

•

For the noise detection and truncation compensation algorithms to work optimally, the impulse response should be sufficiently long to have a late part with a stationary noise floor, also after filtering in the analysis bands.

•

Observe normal caution required for high quality MLS measurements, like avoiding time wrapping, distortion and long averaging times in slowly time-varying systems.

•

It can be worthwhile spending some time finding the right setup and level settings when measuring in a new room, to find the optimum trade-off between background noise (acoustical and electronic) and spurious components due to distortion. Loud is not always best in terms of resulting signal-to-noise ratio, due to loudspeaker distortion.

•

Always control the measurement quality while there is still time to repeat the measurement. Check normalized, logarithmic impulse responses to check whether the response is sufficiently raised above the noise floor and distortion components. Check room acoustical parameters for sensibility. Check decay ranges. Check filtered Schroeder plots to pinpoint anomalies.

•

The calculations will in general be meaningless if performed on non-impulsive sound records, even if results may be shown.

•

The calculations might be inaccurate when performed on impulse responses with very short decay times, for instance electronic systems (amplifiers, filters, etc.). The decay curve formation is based on the assumption that decay times are on the order of at least a tenth of a second, like it will be in any room but an anechoic chamber.