Report of the meeting of the SC-04-08 Working Goup on Measurement and Equalization of sound systems in rooms, of the SC-04 Subcommittee on Acoustics, held in Warsaw, Poland, 2015-05-09
The meeting was convened by interim chair Eddy B. Brixen.
The secretary presented the formal notice on patent policy.
The agenda and report of the previous meeting, held in Los Angeles 2014-10-10, were approved as written.
Development Projects
AES-X215 - "Liaison with SMPTE project to codify current procedures to calibrate the cinema B-chain"
Since the last meeting, the AES comments have been reported to the SMPTE. This report fulfills the task and the project is now closed.
AES-X216 - "Liaison with SMPTE project on Calibration Pink-Noise Standard and Test File"
Since the last meeting, the AES comments have been reported to the SMPTE. This report fulfills the task and the project is now closed.
AES-X218 - "Measurement and calibration of sound systems in rooms"
The main topic of the meeting was this project.
One inspiration for this project is, of course, the SMPTE document. However, as this solely apply to digital cinema sound systems; it will not completely cover the needs of the AES-project, as X218 should be relevant also in connection with rooms that are more complex and complex systems. The implication of amplified microphones is one issue.
Number of microphone positions or measurement points The standard deviation more or less defines the number of measurement points. Brian Vessa mentioned that his upcoming report regarding cinema check (to be finished by September 2015) suggested five positions. This procedure is not for the calibration of new installations, just for the verification of existing ones. Further, it primarily deals with rooms having only one seating area. For venues having more than a primary seating area, for instance, rooms with balconies, this has to be evaluated separately.
Cengale and Mateos from Dolby presented a paper in 2014. The title is "Effect of microphone number and positioning on the average of frequency responses in cinema calibration" (AES Preprint 9083). This paper concludes that there is no reason for more than 8 microphone positions within a single area.
In rooms with less homogenous sound distribution it is not possible just to average eight microphones, as some may happen to be positioned in poorly covered spots. Each point should be evaluated separately. Averaging data is throwing data away. Bad systems may require more microphone positions. However, more measurements does not make the speaker system perform better. To avoid an unrealistic high number of positions, rooms should be classified so the expectations more or less define how detailed measurements that should be performed.
Among the members of the group, the implementation of 8-9 microphone positions seem to be common and scaling up for larger rooms.
Thomas Lago: Taking the coherence into account gives the possibility of knowing whether the measurement relates to the speaker measured. The accepted coherence can be defined in relation to the quality of the measurement. However, in rooms like concert halls for classic music the coherence may become low due to reverberation. If basing measurements on coherence the limits must be defined depending on room type.
The group shortly touched on measurement techniques for mapping measurement results, including the involvement of wireless techniques.
What kind of measurement microphone should apply? Diffuse field / free field / Head and Torso? If the coherence is low it doesn't matter what kind of microphone or is chosen. On the other hand, HATS provide a result closer to the perceived sound although they are expensive and time consuming to move around. Less expensive ear-simulators should be available but still, implementation is not a feasible solution at this point. In a near future it should be considered to apply microphone arrays for 3D measurements as it is already implemented for concert hall measurements. Object based audio also apply ceiling speakers, thus a 3D presentation of the sound field can be of interest. However, so far we are aiming for the simple solution utilizing single point pressure microphones.
No final conclusion was made regarding the number of microphone positions. However, all group members were requested to forward ideas on best practice based on their daily work.
Measurement signal:
It should be possible to synchronize the measurement system to the signal. This excludes the old method of just applying a pink noise generator and do the measurement with a RTA. Therefore, the final methodology should build on impulse response by applying MLS or the like.
Front-end microphones / acoustic feedback:
The basic purpose of many PA/SR systems is to provide amplification to microphone signals. Too much amplification produces unwanted acoustic feedback. Do we need to develop a method to measure SPL before feedback in order to document the threshold of instability? The NT ACOU 108 has a method that provides data on this. In the discussion, it was mentioned that it should be possible to document this basic system. However, there are many parameters to define like microphone position, microphone directivity and so on. No conclusion was made on this topic. The group should look up papers related to acoustic feedback problems.
Speech Intelligibility:
On this, we rely completely on IEC 60268-16 (Speech Transmission Index). The measurements are carried out in exactly the same microphone positions as applied for other measures.
Maximum SPL:
It should be possible to document the maximum SPL of a system. In some cases like in the calibrated cinema, the required level is given. In other systems is must be defined by some kind of non-destructive procedure. We need to define for how long time this max SPL should be present. Similar considerations exist on the component side. The AES2 standard actually goes into this. The problem is to define short term / long term (Long term = lifetime?). We will learn from the AES2 revision to define common terms if possible.
Acoustical noise:
It is important to measure the emitted acoustic noise from the system (hiss from speakers, ENF-components, fans, etc.). It was suggested to document this by measuring spectra. One measurement of the noise spectrum with the system on and one or two measurements with the system off; two measurements, for worst and best case scenario regarding room settings of ventilation, light and so on.
Distortion:
Distortion is very difficult to measure. Often the rattling of mechanical elements in the room are heard before the distortion. If considering the coherence measurement it is avoided to perform distortion measurements. If the system distorts the coherence simply is reduced. Actually, this is a topic for further study. We should encourage universities to let students study this for their thesis.
Presenting data:
We should recommend standard forms for the reporting of measured results. It should not be mandatory; however, in most cases it would ensure full and sufficient information for a report. The SMPTE standard will include a number of forms. We will get inspired from this when published.
Liaisons
There were no new liaison issues.
New Projects
Bob Lee and Jan Abildgaard have mentioned a project: Amplifier sizing for passive loudspeakers. It is not yet an official project but we can expect this to end up in SC-04-08.
John Woodgate asked for a test method for rating human speakers with regard to achieved speech intelligibility. No method exists - but needed.
New Business
There was no new business.
The next meeting will be scheduled in conjunction with the 139th convention in New York, NY., US., 2015-10.
